This week on the GeekWire Podcast: It was a big week for cybersecurity for Seattle’s cloud giants, albeit in very different ways for each.

Microsoft President Brad Smith was in Washington, D.C., testifying before the U.S. House Homeland Security Committee about Microsoft’s security challenges — stay tuned for highlights at the end of the show.

Amazon held its annual AWS re:Inforce cloud security conference in Philadelphia. Generative AI has added some big new wrinkles to cybersecurity, and that was one of the main topics in my recent conversation with one of the people who keynoted the AWS event this week, Steve Schmidt, Amazon’s chief security officer.

Listen below, and continue reading for highlights, edited for context and clarity.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

How generative AI is changing the security landscape: “Generative AI very definitely does enable attackers to be more effective in some areas. For example, when crafting more effective phishing emails, or crafting solicitations for people to click on links or things like that, it definitely enables the attacker a lot more.

“But it can also enable the defender, because when we take advantage of generative AI, it allows our security engineering staff to be more effective. It allows us to unload a lot of the undifferentiated heavy lifting that the engineers had to do before, and to let them do that thing that humans are the best at, which is looking at the murky gray area, and sifting through the little tiny pieces that don’t seem to make sense, and putting them together in a puzzle picture that all of a sudden makes them go, ‘Aha. Alright. I know what’s going on here.’

“In most cases, when we apply generative AI to the security work that we have to do, we end up with happier security engineers on the other end, because ultimately, they don’t want to do the boring, laborious stuff. They want to apply their minds. They want to think about the interesting angles to the stuff that they’re working on — the stuff that generative AI can’t do right now.”

One use case for generative AI in security: “An easy example is plain language summarization of very complex events. If you think about the security job that I’ve got here, a lot of it is taking little tiny pieces of technical data and forming them into a story about what’s going on.

“Creating that story, and then taking that information and conveying it to business owners, is something that every security professional has to do. It is arguably one of the hardest parts of our job — taking something that’s incredibly complex, technical and nuanced, and putting it in a language that makes sense to a chief financial officer, or a chief executive officer. Generative AI is actually turning out to be very useful in that space.”

Three big questions companies should ask to adopt generative AI securely:

1. Where is our data? “Business teams are sending data to an LLM for processing, either for training or to help build and customize the model, or through queries when they use that model. How has that data been handled throughout that workflow? How was it secured? Those are critical things to understand.”
2. What happens with my query, and any associated data? “Training data isn’t the only sensitive data set you need to be concerned about when users start to embrace generative AI and LLMs. So if your user queries an AI application, is the output from that query and the user’s reaction to the results used to train the model further? What about the file that that user submitted as part of the query?”
3. Is the output from these models accurate enough? “The quality of the outputs from these models is steadily improving. And security teams can use generative AI as one of the tools to address challenges. From the security perspective, it’s really the use case that defines the relative risk.”

How Schmidt’s past experience at the FBI informs his approach: “The thing that I took the most out of my experience at the FBI was a focus on the people behind adverse actions. A lot of my career, for example, I was focused on Russian and Chinese counterintelligence. If you look at the motivators for espionage, in the classic world, they’re exactly the same things that are motivators for hackers, right now. It’s money, ideology, coercion, or ego.”

What he gets from his volunteer work as an EMT and firefighter: “As people, we crave feedback. We want to see that we are successful, we want to see that what we do matters. And in the computer world, a lot of what we’re dealing with is virtual. So it’s really hard to see the result of your action. It’s also really hard to see an individual impact in an area where you’re looking at, like I am, hundreds of millions of machines.

“Being a volunteer firefighter, and advanced Emergency Medical Technician, means that if I do my job well, an individual human being who I can see and touch has a better day. And I get that real human feedback that isn’t available from a computer. That’s incredibly satisfying. As a person, I know, I am personally bringing value to this; I am helping that person in a situation which may have been the worst day of their lives, and we’re going to make it better.”

Listen to the full conversation above, or subscribe to the GeekWire Podcast in Apple Podcasts, Spotify, or wherever you listen.

Audio editing by Curt Milton.

“You need to use your human intelligence. No AI will take this away from you. … You need to just realize how beautiful this is, to be a human. And you will see that you can do so many more things, thanks to this [technology].”

Our guest this week on the GeekWire Podcast is Léonard Boussioux, an assistant professor in the Department of Information Systems and Operations Management at the University of Washington’s Foster School of Business, and adjunct assistant professor at the UW’s Allen School of Computer Science and Engineering.

Boussioux received his doctorate in operations research from the Massachusetts Institute of Technology. His research combines areas including machine learning and AI with a focus on healthcare and sustainability. Last year, he launched a class called “Generative AI in the Era of Cloud Computing” at the Foster School.

Related links and articles:

• Boussioux’s website, including photography and AI art: www.leobix.us
• Foster School: A New Era Of AI And Human Ingenuity In The Classroom
• TEDxMIT Salon presentation: Let Art Be Your Superpower
• TEDxBoston: How to Solve the UN’s Sustainable Development Goals in 5 mins.
• GeekWire: How a photo of OpenAI’s Sam Altman, enhanced by AI, sparked a journalistic debate at GeekWire
• Wall Street Journal: The AI Revolution Is Already Losing Steam

Listen above, and continue reading for highlights from Boussioux’s comments, edited for context and clarity. Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

The fundamental appeal of AI: What I like a lot about AI is the fact that it bridges gaps. I like that it’s making everything more multidisciplinary. It used to be that everything was very siloed. You studied chemistry, you studied physics, pure math. I believe that AI can help people come together and work together.

I also love that AI is able to upskill people. I think that AI, because it has some good knowledge in many different topics, can really help people do things that were deemed unfeasible before.

For instance, my business school students, typically they rarely code, if at all. Now, I taught them in just 30 minutes how they can build a website from scratch in coding languages like HTML and CSS. Many of them did not even know those two acronyms. And they all managed to build a beautiful website, thanks to the power of AI.

So this is really the beauty of helping people to be more creative, but also for creative problem-solving.

Using AI to tap into our artistic selves: What I believe in is the fact that us humans have something beautiful. It’s our capability to connect with each other, to build communities, to create. And I believe the best way to create is to tap into our artistic selves.

Our society is not always emphasizing how important this is, to be creative, or that we are all artists. Very often people decide that, being an artist is not for me. I don’t have the time. I don’t have the skills. I believe that, no, every one of us can be artists. And I also believe that AI is an opportunity to help us become more artistic in our daily lives.

On the idea that AI advances are losing steam: I believe, honestly, that AI is going for now in a linear trend. Between GPT 3.5 and 4, it was a very nice improvement. GPT 4o has seen some nice multimodal improvements, nothing much in terms of capabilities of the model.

The exponential part is not necessarily in the tool, but in the humans. You need to use your human intelligence. No AI will take this away from you. You won’t be replaced anytime soon. …

You will need to be creative. You will need to be an artist, to think out of the box, to figure out those little details that nobody else will see. And this is an opportunity to leverage the technology to get you [there] faster or differently or get the right support you need. But you still need to use your brain, ultimately. You need to just realize how beautiful it is to be a human. And then you will see that you can do many more things, thanks to this.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Audio editing by Curt Milton.

This week on the GeekWire Podcast, we explore the frontier of crowd-augmented cognition, the concept of humans working together with the help of technology, including new ways that artificial intelligence is changing the field.

Our guest is Aniket (Niki) Kittur, a professor in the Human-Computer Interaction Institute at Carnegie Mellon University, where his research focuses on new methods of augmenting human intellect using crowds and computation.

We also talk about a related project that Kittur and his colleagues developed called Skeema, a browser tab manager that helped users organize their work, projects, and ultimately their brains in the process.

Listen below, or subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen. Continue reading for highlights from Kittur’s comments, edited for context and clarity.

The focus of Kittur’s research: My lab is really interested in how we help people make sense of overwhelming information. Online, we’re constantly bombarded with all sorts of information that we have to make sense of.

We’re looking at how we take what people are good at, and what machines are good at, and put those together to help us understand, make decisions, and be creative with that information, better than we can alone.

Studying crowdsourcing movements like Wikipedia and Mechanical Turk: We’ve been looking at these very different ways of how we can put people together to go beyond what they’re capable of alone, and how you create these architectures that combine people with machines, and the right ways to do that.

AI and human creativity: We have one project, working with Toyota, to help their automotive designers be more creative … and we’re using AI there to help them find inspirations from very different fields, such as how a crow flaps its wings and creates vortices that stabilize its flight path. Can you use that for better mobility in different situations? We use AI to unlock those things that are hard for people to think of, and then pull those into the domain that the designer is trying to solve.

How the Skeema browser tab organizer fits into the lab’s work: The vision in my work is, how we create a force in the world that can start to stitch information back together, stitch it back into knowledge, so that we can constantly be improving and learning faster, being more innovative, solving the problems we need to solve.

Skeema is one step in that direction. Skeema right now is a very individual thing, where you’re taking the fragmented information in your browser and trying to pull it together and make it useful for yourself. But imagine that we can start to connect people who are pulling this together, and [help them] build on other things.

Large language models and the human brain: If you look at the brain, it’s not made up of just one type of thing. It’s actually a collection of lots of different components. Our memory works in a certain way, it’s super-parallel, but our ability to reason is super-serial, and can only hold a few things. And we put those together, along with vision and all these other modules that work in different ways.

So why don’t we think about this as, at a larger scale, an information-processing problem? And LLMs are definitely one of those that have different characteristics that we can put together to help us with our human goals.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Audio editing and production by Curt Milton.

Microsoft and major PC makers this week announced a plan to incorporate artificial intelligence directly into personal computers, adding on-board neural processing units (NPUs) as part of a new architecture that promises better performance, longer battery life, and local AI processing that unlocks new capabilities and features.

We discuss the new Copilot+ PCs with our guest on this week’s GeekWire Podcast, Stefan Weitz, an investor and entrepreneur who worked at Microsoft for 18 years in groups including Microsoft Bing and MSN. He is the founder of the new HumanX conference on AI taking place in March 2025 in Las Vegas.

Weitz is bullish on the potential for Copilot+ PCs to improve the overall computing experience and motivate Windows PC users to upgrade their machines, a long-awaited milestone for Microsoft’s flagship operating system.

The new Copilot+ PCs have also raised security and privacy concerns, focused primarily on the “Recall” feature that takes regular screenshots of user activity on the machine, creating an index that can be queried using AI.

The company released an FAQ that emphasized the underlying security and privacy controls for users, but also made it clear that the feature will be activated if users accept the defaults during initial bootup of a new machine.

Weitz also shares some of his favorite AI tools, including Cleft Notes and Read AI.

Listen to the episode above, and subscribe to the GeekWire Podcast in Apple Podcasts, Spotify, or wherever you listen.

Audio editing by Curt Milton.

[Editor’s Note: This week’s GeekWire Podcast features an interview with author Loretta Napoleoni, conducted by Ross Reynolds, the longtime public radio host and occasional GeekWire Podcast guest host, in an event presented by Town Hall Seattle on April 18.]

Loretta Napoleoni is an Italian economist and journalist whose books include “Rogue Economics” and “Insurgent Iraq.” Her newest book is “Technocapitalism: The Rise of the New Robber Barons and the Fight for the Common Good.“

In this interview, she explains the concept of the “present future.” This refers to the disorienting era in which we’re living, characterized by rapid technological change that creates anxiety and a feeling of being constantly behind.

She also discusses the control of technology by a few powerful entrepreneurs, whom she calls “technocapitalists,” and the failure of society and the state to direct technological innovation for the common good.

Listen below, or subscribe to GeekWire in Apple Podcasts, Spotify or wherever you listen. Continue reading for highlights from Napoleoni’s comments.

The rapid pace of technological change: “The ‘present future’ is a central concept of the book because, in reality, what it means is that we’re living in the present, of course, but we’re already living in the future. And this is because technology moves so fast that it is almost unbearable for the human being to keep pace with this speed, with the speed of change, especially technological innovation. And that creates anxiety because you’re constantly feeling that you are behind, that you are not adequate to the modernity of life.”

The rise of technocapitalism: “This is not a critique of capitalism. It’s more a critique of how society and the state have approached the control in the hands of the tech titans. We were unable to direct technological innovation towards the common good, towards the society as a whole. I don’t want to sound like I am against them. I don’t want to sound that they are the bad guys and we are the good guys. What I’m saying is, it’s a failure of the system.”

Government’s inability to keep pace: “I don’t think the government wants to catch up. That’s the problem. The EU tried to regulate certain aspects of the internet. There is privacy legislation which was introduced in Europe, which is actually very, very good. The Americans didn’t. The Americans were very much against it. Things have changed recently, but have changed because we reached a certain level in which you do need to intervene. But it is too late.”

The crux of the problem: “Everything that is happening is happening because of political choices. And the political choices are related — because we live in a democracy — to our decisions. We are the voters, and we put those people in place and those people do certain things. So at the end of the day, it’s not technology that’s the problem. The problem is us.”

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Thanks to Town Hall Seattle for presenting this event and providing today’s audio.

This week on the GeekWire Podcast, we go inside the GeekWire Awards, our annual event recognizing the top leaders, companies, and innovations in the Pacific Northwest, which drew a sold-out crowd to the Showbox SoDo in Seattle.

We talk with finalists about what’s next, the economy and key trends in their industries, and we hear from some of the winners on stage during the program.

Guests on the show include:

• Gaurav Oberoi, the CEO of Lexion, the AI-powered contract management software company, which this week announced an agreement to be acquired by Docusign for $165 million.
• Linda Lian, CEO of customer intelligence platform Common Room, a past Startup of the Year in the GeekWire Awards. Lian was a finalist for CEO of the Year in this year’s GeekWire Awards.
• Gordon Pan, the president of Baird Capital, the private equity and VC firm that is part of the Baird financial services and wealth management firm.
• Todd Owens, CEO of Kevala, a finalist for Startup of the Year, which helps senior living and healthcare organizations manage scheduling.
• Steve Helmbrecht, CEO of Treasury4, a finalist for Deal of the Year, which raised $20 million last fall to help manage financial and treasury processes for enterprises and public sector professionals.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

RELATED COVERAGE

• Photos: Inside the GeekWire Awards, with the big winners, musical numbers and surprise guests
• GeekWire Awards 2024 revealed: Winners bask in momentum of AI and potential of region
• Hot number at the GeekWire Awards: Watch ‘The Drunken Tenor’ light up tech from bygone era

More: GeekWire Awards 2024

This week on the GeekWire Podcast, we discuss Amazon’s efforts to compete with Trader Joe’s, and the related tactics revealed in “The Everything War,” the new book by Dana Mattioli of the Wall Street Journal. We also revisit last week’s episode with Mattioli and share some of the reactions to the discussion.

Plus, the FTC probes Amazon’s internal use of Signal’s disappearing messages feature, a newly disclosed email shows how Microsoft scrambled to catch up in artificial intelligence, and a new report says that Bill Gates is still a highly influential figure inside Microsoft, especially when it comes to the company’s AI strategy.

Finally, we go inside the Seattle courtroom where a prominent figure from the cryptocurrency world, Binance founder Changpeng “CZ” Zhao, received a controversial prison sentence this week.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Other stories discussed on the show:

• Amazon stock rises as Q1 earnings top estimates with $143.3B in revenue; AWS sales up 17%
• DoorDash rips Seattle over minimum wage law; reports record quarterly revenue of $2.5B
• Prudential to shut down Assurance, the insurance tech startup it acquired for $2.35B in 2019
• KING5: Binance founder Changpeng Zhao sentenced to 4 months for allowing money laundering
• Business Insider: Bill Gates is still pulling strings at Microsoft, overseeing AI ideas
• Internet Archive: Full text: An epic Bill Gates rant

This week on the GeekWire Podcast: A conversation with Dana Mattioli, Wall Street Journal reporter and author of the new book, “The Everything War: Amazon’s Ruthless Quest to Own the World and Remake Corporate Power.”

Mattioli spoke with more than 600 people for the book, including current and former senior Amazon leaders who were interviewed without the company’s knowledge. The book examines Amazon’s approach to expanding its dominance, the culture established by Jeff Bezos and other senior leaders, and the practices that led to the Federal Trade Commission’s antitrust suit against the company.

Amazon disputes a number of the book’s assertions, as detailed in the episode.

This week brought a new twist in the FTC lawsuit, as the agency asked a federal judge to order Amazon to disclose details about internal use of Signal’s disappearing messages by Bezos, Amazon CEO Andy Jassy, and others. Amazon called the agency’s assertions baseless, saying it has been forthcoming on the issue.

Listen below, and continue reading for excerpts from Mattioli’s comments, edited for clarity and length, along with Amazon’s responses to GeekWire about specific issues raised by the book, after it was published this week.

The process of reporting the book: “I spent three years reporting it. We had hundreds of pages of internal documents. So many sources spoke to me, and I tried to get as many of them as I could on the record. It’s a damning look inside Amazon’s business practices. And I knew it needed to be bulletproof in terms of my sourcing.”

Amazon’s lax internal controls for confidential data: “I learned that Amazon, even though they’re one of the biggest, most sophisticated companies in the world, has very primitive firewalls in place to protect confidential data.

“That’s not limited to private-label. [Amazon teams have allegedly used third-party seller data to advantage the company’s own products.] I heard this in other areas of the company, as well, especially its venture capital arm.

“There’s an anecdote in the book where a founder who’s meeting with the Alexa Fund for an investment sets up a data room, and she puts all of her proprietary information in that data room. Amazon’s also her competitor, so it’s an uncomfortable situation. She learned almost immediately that there were people throughout the Amazon organization, not affiliated with the Alexa Fund team, looking at her proprietary documents.

“As one employee told me, ‘The company still had operated in this really startup-y way, meaning you often had access to data for your job that you didn’t need access to. When you’re moving fast, you’re not really thinking clearly about, “Well, how do we make sure that only the right people have access to this dataset?” The internal infrastructure of the company is basically popsicle sticks, and like duct tape.’ “

Amazon’s response: “We do not condone the misuse of proprietary confidential information, and thoroughly investigate any reports of employees doing so and take action, which may include termination. We have strict policies in place and keeping proprietary information secure is embedded into how the team operates.”

“We use publicly available data to inform our strategy in service of providing the best experience for our customers. But we draw a clear line against using non-public, seller-specific data to compete with sellers, and our policy goes further than any retailer we know of.

“We take this policy seriously — we thoroughly train on it, audit, and investigate any reports of violations. Public allegations of policy violations have been based on misunderstandings and not on credible evidence. Any suggestion that there was pressure from Amazon executives to violate this policy in developing our private brands is flat out incorrect.”

The role of Amazon’s culture: “I’ve covered companies for nearly 18 years at the Wall Street Journal. This culture is so different from any company I’ve ever covered. I think some of the senior leaders aren’t aware of how much pressure this environment creates for their employees.

“That’s why it reminded me of Wells Fargo. It wasn’t like the CEO of Wells Fargo was saying, ‘Open up all these accounts; don’t tell the clients.’

“Some people describe it a little bit like ‘The Hunger Games,’ where you have to perform, and you’re competing against equally brilliant employees every day, and none of you wants to be in that bottom 6%. None of you wants to risk getting cut before you get to your restricted stock units. It does incentivize some people to do things that they’re not supposed to do.

“Jeff Bezos is not directly telling them to do that, of course, but it happened.”

Amazon’s response on this issue, which was also raised in the book: “This is false, and not supported by any material the author has presented. Amazon’s culture centers on innovating for customers to make their lives better and easier.”

How social media posts by Amazon execs undermined the company’s public policy work: “The D.C. team is trying to mend bridges and make connections on the Hill and thwart litigation that could break Amazon up or hurt it.

“They had a process called ‘watering the flowers’ where they were trying to make friends and influence people on the Hill. And then Jeff and his team would often come in and just stomp all over those flowers, and send mean tweets to people that this team was trying to cozy up to, and it just undermined the efforts.”

Parallels to Standard Oil: “Standard Oil was also doing things like spying on their competitors. Even though this was not the digital age, they had near-perfect intel on the oil refining environment because of how big they were. And that’s a theme that comes up in my reporting on Amazon: Private label, Alexa Voice Services, the Alexa Fund.

“One of the tactics that Standard Oil was also criticized for was predatory pricing, and forcing competitors to sell to them under the threat that, if you don’t, then we’ll put you out of business.

“There’s a scene in the book where Amazon threatens predatory pricing toward Diapers.com, and says, if you sell to someone else, we’ll cut the price of our diapers to zero, which brings it to its knees. So there are some parallels.”

Amazon’s response: “Standard Oil Trust controlled roughly 90% of the refined oil in the United States. Amazon accounts for about 4% of U.S. retail. Any comparison of the two companies is not based in reality.”

[Editor’s Note: Amazon represents about 40% of US ecommerce sales.]

Amazon’s impact on prices: “The FTC is alleging that, because sellers know they need to be on Amazon — 40% of all ecommerce in the U.S. happens there — that Amazon is able to levy all sorts of fees on them. And that’s exploded in recent years. … As a result of that, sellers have had to raise their prices for consumers to offset all the fees.”

Amazon’s response: “The company fundamentally disagrees with the FTC’s allegations because they are wrong or misleading, and they would harm consumers and independent businesses.”

Mattioli’s motivation for writing the book: “What I hope this book does is that it lifts the veil on a purposefully secretive company, that it reveals to readers just how Amazon wins all the time, and shows instances of how Amazon has its fingers on the scale. And I think we achieved that.”

Amazon’s response: “Amazon’s success is the result of continually innovating for consumers and small businesses over three decades to make their lives better and easier every day. The facts show Amazon has made shopping easier and more convenient for customers, spurred lower prices, enabled millions of successful small businesses, and significantly increased competition in retail.”

“The Everything War: Amazon’s Ruthless Quest to Own the World and Remake Corporate Power,” by Dana Mattioli, was published by Little Brown on April 23.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

This week on the show, we offer a sneak preview of the GeekWire Awards, at the Showbox SoDo in Seattle on Thursday, May 9.

• We look back at the past winners of the Next Tech Titan title in the awards, and consider the event’s track record in predicting major companies to emerge from Seattle and the Pacific Northwest.
• We also contemplate the evolution of the “Workplace of the Year” category through the rise of remote and hybrid work.
• And finally, we consider the impact of migration on another part of state: Spokane, Wash., and the Inland Northwest.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

See more GeekWire Awards coverage, and learn more about the event, which is presented by Astound Business Solutions.

Audio editing by Curt Milton.

Early versions of generative AI were “like a dancing bear,” says Werner Vogels, the Amazon chief technology officer. “We’re all amazed that this bear can dance,” he explains. “We don’t really look at whether it dances well or not.”

In the past two years, with ample training, the bear has become a better dancer.

However, he says, “this is still technology that relies on you, as an individual, actually taking action based on what this technology tells you. It’s a tool. Artificial intelligence makes predictions. We as humans take actions based on that.”

That’s one of the highlights from our conversation with Vogels on this week’s GeekWire Podcast, recorded at Amazon in Seattle. We talk about the evolution of AI, its promise in healthcare and the environment, his broader tech predictions for the year, and one of his most iconic traditions at Amazon’s annual cloud conference.

Listen below, and keep reading for excerpts, edited for context and clarity.

The rapid evolution of technology, and the changing nature of education: Technology has been evolving exponentially fast. Let’s say you have a computer science degree. You probably weren’t exposed to large language models during your education. So there’s a lot of on-the-job training that still needs to happen. That’s a shift that is happening, where the university four-year education is much more about critical thinking, and learning how to learn, because technology will be a lifelong learning experience for anyone. … Individual companies like Amazon will be doing the education themselves, assisted by these large language models.

The rise of culturally aware AI models: It has triggered a number of very interesting things. For example, our Japan team, with a whole range of our customers, did a hackathon and built many of these culturally aware Japanese models. These models can compete with each other, and talk to each other, and agree on what the answer should be.

The potential of fine-grained sustainability reporting: I believe we as cloud providers, not just AWS, we need to get to the point where we not only report [to customers] on costs, but report on milligrams of CO2 used by your particular service, for this particular period of time. We’re not terribly far from there. That will definitely happen. At this moment, the amount of resources that you’ve used in general with a pay-as-you-go model means you pay more if you use more resources. So it’s already a pretty good proxy for sustainability. But we need to present it in a way to our customers that is true and accurate, in ways that they can also use it towards their regulatory agencies, for example.

His selection of t-shirts from different bands for his annual AWS re:Invent keynote: I’ve always been a t-shirt and jeans guy. I pick every year something that I think will be fun. … If there is a link between the different t-shirts, it is that I feel that those that I have promoted have been innovators in their type of music.

The broader social responsibility of the technology industry: We as technologists have the responsibility to make sure that our technology is used responsibly, that there are tools next to this technology where we can investigate whether it’s being used correctly. … We need to make sure that we have these large data sets available for everyone to use. Good AI requires good data. Good work requires good people. And we as technologists should be stepping in there to do the right thing for this planet.

Listen to the full conversation above, or subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Related links

• Werner Vogels – All Things Distributed
• Archive of Now Go Build episodes
• AWS re:Invent 2023 – Keynote with Dr. Werner Vogels
• Tech Predictions for 2024 and beyond
• @Werner on X.

Audio editing by Curt Milton.

A few years ago, Microsoft researchers and data scientists used machine learning to help marine biologists analyze patterns in underwater recordings of beluga whales.

Upon learning about the project, another group asked if it would be possible to use a similar approach to analyze audio from the Syrian war, to detect the use of weapons banned by the Geneva Conventions. The answer was yes.

That’s the kind of inspiration that Microsoft’s philanthropic AI for Good Lab hopes readers will draw from its new book, “AI for Good, Applications, in Sustainability, Humanitarian Action and Health,” to be published April 9. 

“For us, it is really important to show real-world examples of how we can use AI to solve these problems,” said Juan Lavista Ferres, Microsoft corporate vice president and chief data scientist, who directs the AI for Good Lab. The goal is for other scientists and researchers to see, in these examples, new ways to use AI to solve other societal problems.

The book offers a series of deep dives on the lab’s projects, conducted with outside researchers, non-governmental organizations, and other experts.

The case studies show the potential of AI to do good in the world, but they also give a clear-eyed and practical look at the risks and limitations.

“We need to make sure that we understand the data that we’re using,” Lavista Ferres said. “This is, for us, why it’s so critical to work with subject-matter experts, to better understand the problems that we’re trying to solve.”

In advance of the book’s release, Lavista Ferres joins us on this episode of the GeekWire Podcast for a conversation about the lab’s work, the potential for AI to bring about positive change in the world, and takeaways for the rest of us as we look to apply artificial intelligence to our daily work and lives.

Listen above, or subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen. Continue reading for highlights, edited for clarity and length.

The potential for AI to do good in the world: “Since the inception of AI, the broad majority of the use cases for AI have significantly helped society, and we expect these to even be broader. With more capabilities, including large language models, we’re now solving problems that before we couldn’t solve. I’m extremely optimistic about the uses of AI. … I think that we’re just starting to see the type of impact we can get from AI, and I hope in the next 5-10 years, we will see much more of the impact that this technology can have on society.”

One of the keys to a successful AI for Good project: “In order for us to have an impact, we need people to be using these models in a production setting. There’s a big difference between solving a problem in theory and solving a problem in practice. It’s important to work with organizations that have not only the subject-matter expertise, but also the capabilities to put these models into production. Our job is to help them solve these problems. But their job will be to eventually work with these models in production settings. And that’s challenging.”

The lab’s work with the Carter Center: “The problem that they’re trying to solve is to have a near real-time assessment of potential conflicts around the world. They get this information from news sources, sometimes in multiple languages, across the world. Usually they would have many analysts reading this information and trying to come up with with an assessment. We took that dataset that they’ve been using, and we used natural language processing. … These models were as good as these experts on classifying whether there was a conflict or not, or what type of conflict it was, which allowed these experts to be focused on what they do best, trying to see what they could do with that data. So that has been really impactful. They have all of this in production, and we continue to collaborate with them.”

The critical role of data, and the role of open datasets: “People say data is the new oil. Data is the new code. It is clearly the most important portion of an AI model. It is really important to make sure that we don’t introduce biases in the way that we collect the data. But there’s also a significant amount of open datasets out there. And that’s something that we also contribute to society whenever we work on these projects, when possible. … We’re not alone. There’s a whole open data movement where more and more organizations are open-sourcing their datasets.”

What’s next in the adoption of AI for good: “We will see many more tools that will allow anybody to do things that before would require coding skills, and we are seeing that already. We expect this trend to continue. More and more, it will become easier for people to use this type of technology to solve their problems. … As a society, we have a responsibility to maximize the use of tools, and minimize any potential use of this technology as a weapon. It’s critical for society to work together to make sure that this technology can be used for good.”

“AI for Good: Applications, in Sustainability, Humanitarian Action and Health,” by Juan M. Lavista Ferres, William B. Weeks and researchers from Microsoft’s AI for Good Lab, will published by Wiley on April 9, 2024.

Audio editing and production by Curt Milton.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

This week on the GeekWire Podcast: Reporter Kurt Schlosser takes us inside a Tesla Cybertruck, explains what it was like to drive one, and tells us how people reacted to the futuristic vehicle as it debuted on the streets of Seattle.

Also on the show, Kurt shares details from his recent story about Nala, the Seattle Police Department’s electronics-detecting police dog, and explains how these law-enforcement K9s are trained to find all sorts of devices.

And in our final segment, we hear a clip from comedian Pete Ballmer, the son of Steve Ballmer, about growing up as one of the kids of the former Microsoft CEO.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

This week on the GeekWire Podcast: We peel back the layers of Microsoft’s hiring of AI pioneer Mustafa Suleyman and colleagues from Inflection AI, considering how it plays into the trend of the Redmond company striking deals that look a lot like acquisitions, without actually being acquisitions.

Then, we delve into the U.S. Justice Department’s landmark antitrust lawsuit against Apple, and consider the parallels to the DOJ’s prior case against Microsoft.

And finally, we ponder Washington state’s plan to fight graffiti with drones.

Additional Links

• S. “Soma” Somasegar: Analysis: Microsoft’s big hires foreshadow bold new move into consumer AI
• Newcomer: Microsoft’s Non-Acquisition Acquisition (second item)
• The Information: Microsoft Agreed to Pay Inflection $650 Million While Hiring Its Staff
• Bill Barr in the Wall Street Journal: Siri, Does Apple Violate Antitrust Law?
• DOJ: Justice Department Sues Apple for Monopolizing Smartphone Markets
• The Verge: DOJ’s sweeping Apple lawsuit draws expert praise

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

This week on the GeekWire Podcast: The rise, fall, and revival of Zulily.

We revisit one the most prominent e-commerce brands to come out of Seattle, explain its decline, and consider a plan by Beyond Inc., led by investor and entrepreneur Marcus Lemonis, to acquire its brand assets and relaunch the site.

Zulily was founded in 2010 by two former Blue Nile executives with a focus on serving moms and kids through its online flash sales platform, allowing customers to purchase limited-time deals while using a unique inventory model. This niche focus and sales strategy led to rapid growth, with customer numbers increasing exponentially in its first few years.

It grew quickly and went public in 2013 with a multi-billion dollar valuation.

However, Zulily struggled to maintain its hyper-growth trajectory after going public, and its stock price began to fall in subsequent years. It dealt with complaints about slow shipping and competition from Amazon, among other e-commerce retailers.

Zulily was acquired in 2015 by QVC’s parent company for $2.4 billion, but struggled to regain the magic that helped propel the business in its earlier years.

Under ownership of private equity firm Regent, which acquired Zulily in May 2023, the company saw rapid deterioration including layoffs, unpaid vendors, and ultimately liquidation in late 2023, despite still generating hundreds of millions in sales that year.

Lemonis, known for reviving struggling companies on his former CNBC TV show “The Profit,” joined the board of Beyond in 2023 and last month became its executive chairman. Beyond is the parent company of Overstock.com and acquired bankrupt Bed Bath & Beyond’s assets last year.

Earlier this month Beyond announced an agreement to acquire Zulily’s brand assets for $4.5 million.

In an interview with GeekWire shortly after the announcement, Lemonis emphasized returning Zulily to its flash sales roots to drive revenue without additional fixed costs. He also aims to compete more directly with retailers like Wayfair.

Unanswered questions remain around issues including the potential impact on Zulily’s pending antitrust lawsuit against Amazon. E-commerce growth also faces headwinds as the pandemic boom fades and consumer spending slows. However, Lemonis’ involvement raises hopes, and Beyond’s stock is up 80% in the past month.

Artificial intelligence is a powerful technology that promises to reshape the future, but it also poses many challenges and risks. One of the most pressing issues is the lack of regulation and oversight of the data used to train AI models. A new nonprofit, the Seattle-based Transparency Coalition, is aiming to address this issue.

The co-founders of the group, veteran startup founders and technology leaders Rob Eleveld and Jai Jaisimha, join us on this episode of the GeekWire Podcast to discuss their reasons for starting the organization, and their goals to help shape emerging legislation and public policy in this area.

Listen below, and continue reading for notes on the conversation.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Origins and mission: The Transparency Coalition started with a literal fireside chat. Jaisimha and Eleveld discussed concerns about issues with AI transparency and unconstrained training data while camping on Whidbey Island.

They decided to found the Transparency Coalition as a nonprofit organization to address these problems through policy advocacy and public education. Their goal is to promote more ethical and responsible development of AI by increasing transparency into how models are trained, and the data used to train them.

Both have extensive experience as technology and startup leaders:

• Eleveld, a former U.S. Navy submarine officer, was CEO of Ekata, an identity verification company acquired by Mastercard in 2021, after earlier leadership roles at companies including Whitepages, Optify and Shiftboard.
• Jaisimha, who earned his PhD from the University of Washington in electrical and computer engineering, is a UW affiliate professor who worked at companies such as RealNetworks, Amazon, Microsoft and Medio. He founded and led the startup Appnique, which applies machine learning to mobile advertising campaigns.

“I’ve always been a fan of applying AI to constrained problems, well-thought-through data sets,” Jaisimha explained. “And I’d just become concerned about the sloppy nature of data collection practices, and overblown promises about what these algorithms could do. … The heart of it all was the inputs of the AI.”

Their focus right now is two-fold:

1. Influencing state-level policy and legislation through advocacy, testimony, and education of policymakers. They have been actively engaging with legislators in Washington and California.
2. Broad educational efforts to raise awareness and understanding of AI issues among stakeholders like policymakers, business leaders, and the general public concerned with these topics.

Potential implications: Requiring transparency around training data and how models are used could significantly change the scope of AI models. If companies need to disclose what data is used and get consent, the datasets would likely need to be more focused and constrained to avoid using copyrighted or private content without permission.

One effect would be to narrow the scope of AI applications to address specific problems. Transparency could also make the outputs more predictable and accountable since the relationship to the training data would be clear.

“If you have to license training data, it becomes part of your cost of goods,” Eleveld said. “So the projects get narrower and smaller and more focused on detecting Stage 3 pancreatic cancer [for example], as opposed to trying to answer every question ever posed by humanity. We think narrower and more focused generative AI is much better for society. It’s much more controlled. You can trace the outputs … to what the inputs or the training data was.”

Potential legislation could include:

• Standard definitions of key terms like AI, training data, and transparency.
• Requirements for transparency into what data is used to train models.
• An audit mechanism to verify the data used to train the models.
• Ensuring use of personal data and copyrighted content is opt-in rather than opt-out.

Funding: Eleveld said he and his wife are providing the initial seed funding for the Transparency Coalition. It is a 501(c)(4) nonprofit organization, which allows for more flexibility in lobbying and policy advocacy compared to a typical 501(c)(3) charity. They are now seeking grants from foundations, family offices, and others interested in influencing policy, since donations to a 501(c)(4) are not tax deductible like they would be for a 501(c)(3).

Partnerships and next steps: They are collaborating with AI research organizations like the Responsible AI Systems and Experiences group at the University of Washington to help bring forward best practices from researchers. Part of the idea is to connect policymakers with AI thinkers to help address key issues and identify solutions.

“This isn’t just some magic box,” Eleveld said about AI models. “There are inputs and outputs to it like any other system, and it should be broken down and understood at a baseline level. And if it is understood, then people start asking the right kinds of questions, and hopefully coming to some better policy positions.”

Audio editing and production by Curt Milton.

Listen above, or subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

A new report points to a crisis of trust in innovation, and the risk that rapid technological change — especially in the field of artificial intelligence — will fuel increased populism and polarization across societies.

Richard Edelman, CEO of Edelman, discussed these and other findings from the 2024 Edelman Trust Barometer during visits last week with the global communications firm’s clients in tech-heavy Seattle and San Francisco.

His message, as he explained in a blog post this week, was that “acceptance of innovation cannot be taken for granted, that we must spend much more of our time on adaptation and education, not just on R&D.”

• More than 75% of those surveyed for the report expressed trust in the tech industry, while trust in artificial was 25 points lower, at 50%.
• Eight years ago, technology was the leading industry in trust in 90% of the countries the firm studies. Now it is most trusted in half of those countries.
• Trust in AI companies has declined from 61% to 53% in the past five years.

Edelman says this trend illustrates the risks to the technology industry from what he calls its “head-long jump into artificial intelligence.”

GeekWire sat down with Edelman during his visit to the firm’s downtown Seattle office for this episode of the GeekWire Podcast. It was a fitting location, given that the firm was inspired to start the Trust Barometer by the mass protests and riots at the World Trade Organization conference in Seattle nearly 25 years ago.

Listen below, or subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen. Continue reading for edited highlights from Edelman’s comments.

The importance of trust: Trust is the central proposition in a well-functioning economy and society. And it’s four parts: ability, dependability, integrity and purpose. And we find that, after 2008 and the Great Recession, the ability part is almost taken for granted. It’s, “Can you do it all the time? Do you have standards, and do you have a moral compass?” It just isn’t enough to be able to just do. It’s, do you do it well, and you do it consistently?

Public sentiment about AI: People are looking at AI and saying, “We have to do this right.” And when we talk about the idea of suspicion of innovation, innovation should be the greatest thing ever for business. And this should be a golden time for business, because it’s the most trusted.

But if we rush this, if we put it out in a way where government isn’t seen as being able to regulate because it can’t keep up, or if it’s seen as done without context where there’s reskilling or upskilling to take care of people whose jobs are going to end, then we’re going to have a populist reaction.

AI and the 2024 elections: We need to show that, for instance, we’re going to make absolutely certain that the election goes well on information quality. There are 50 elections this year around the world, and AI is going to have a major influence on this. And all the tech companies I know are very concerned about their technology being used well.

What tech companies can do: Implementation, adaptation, and acceptance are just as important as invention. And we are not talking enough about how we’re going to do this well, and the idea that it will only come in its time, when it’s ready. We need to be really clear with people about how the experiments are being conducted.

This is a massive opportunity for business to show that it’s earned its position as the most trusted institution. It should do this revolution with government setting the boundaries, with NGOs on training in the last mile, with media to explain what they’re doing and show that, in fact, this can be done well.

Related links

• Access the full 2024 Edelman Trust Barometer report.
• Edelman: 2024 Edelman Trust Barometer Reveals Innovation has Become a New Risk Factor for Trust
• Richard Edelman: Technology Industry Watch Out, Innovation at Risk
• Axios: Public trust in AI is sinking across the board

At the outset of their new book, “Big Bet Leadership,” authors John Rossman and Kevin McCaffrey identify three “mega forces” that they believe could define the years ahead for businesses and the economy:

1. Disruptive technologies such as generative AI and quantum computing
2. The aging of America’s workforce.
3. Increased spending on entitlement programs such as Social Security and Medicare, and the looming cost of serving the U.S. debt.

They write in the book, “These mega forces will feed into each other like a vortex, building an overriding theme for business and society—that of a chaotic environment of dramatic change with successful business operators realizing productivity and cost model advantages that separate them from their competition.”

Their thesis: companies that thrive in this chaos will be the ones that master the art of “big bets,” transformative moves that dramatically expand their capabilities and the potential of their businesses.

“Big Bet Leadership” is a playbook for systematically making those bets in a way that reduces risk and increases long-term flexibility, drawing lessons from the likes of Microsoft, Amazon, T-Mobile and SpaceX, while also leaning heavily on the first-hand experience of the authors in overseeing and implementing big bets.

Rossman is a former Amazon executive who was a key figure in launching Amazon’s Marketplace, with more than three decades in technology and business innovation in his career. He has authored books on digital innovation, including “The Amazon Way,” advised major companies, and founded Rossman Partners.

McCaffrey has served in strategic roles at companies including T-Mobile and Google, taking part in growth initiatives and transformations including efforts to expand T-Mobile’s “Un-carrier” initiatives beyond its core wireless business. He has also consulted with McKinsey & Co. He now works with Rossman Partners.

Big bets “come by lots of different names: a growth strategy, a market repositioning, a digital transformation, an AI strategy, an operating change,” Rossman explains. “These are all initiatives or strategies that have high potential for impact.”

Why focus on big bets now?

“We’re entering an era where more change is happening, and more disruption is happening,” Rossman says. “The ability to separate the future winners and losers is going to come from the ability to actually make bold moves. Paradoxically, we know that the vast majority of these big bets fail, but yet, we have to do more of them.”

“You can’t wait until you need a big bet to place it,” McCaffrey adds. “If you haven’t already positioned yourself by placing early bets that you keep small … once the industry trends start to become clearer, and the market opportunity becomes clear, it’s too late. Somebody else is going to take the opportunity.”

These are some of the keys to “Big Bet Leadership,” gleaned from the book and the podcast conversation with Rossman and McCaffrey:

• Take a systematic approach, managing a portfolio of smaller experimental bets to gain strategic flexibility, and establishing standard processes like dedicated teams and documentation to maintain momentum on big bets.
• Assign “single threaded leaders,” fully dedicating the best talent and leadership to each big bet through cleared plates and priorities, rather than overseeing big bets as a secondary duty or with divided attention.
• Clear obstacles proactively, gaining executive sponsorship to adjust policies that could inadvertently slow down big bets, such as procurement and IT procedures, before teams are formed and work begins.
• Maintain velocity by prioritizing big bets over day-to-day operations through dedicated resources and teams, as well as metrics to ensure they do not lose momentum through scope creep or competing priorities.
• Develop clear memos and documentation, including a process of critical thinking exercises to fully examine risks, opportunities, and stakeholder concerns before committing to large-scale plans or investments.
• Present three realistic future options, allowing stakeholders to critically assess the tradeoffs of each without being led to a predetermined solution, and surface concerns that teams can address before momentum is lost.
• Learn from failures by systematically deciding to continue investing in strategically important ideas, pivot based on learning, or end initiatives based on outcomes rather than confusion or lack of commitment.

“Big Bet Leadership: Your Transformation Playbook for Winning in the Hyper-Digital Era,” is available on Amazon. Also see John Rossman’s newsletter, “The Digital Leader.”

Audio editing and production by Curt Milton.

Listen to the full conversation above, or subscribe to GeekWire via Apple Podcasts, Spotify, or wherever you listen.

Entrepreneur and investor Chris DeVore had an inside view of Techstars Seattle as one of the original leaders of the local startup accelerator in 2010, before serving as its managing director from 2014 to 2019.

His Feb. 21 post “What went wrong at Techstars,” looked closely at the broader organization’s evolution — including its increased focus on corporate sponsorships and shift to centralized fundraising — as the backdrop for the news last week that Techstars is closing its Seattle accelerator as part of a larger reset also impacting the original Techstars accelerator in Boulder, Colo.

So where should Seattle go from here? And what role do startup accelerators serve in the age of AI and remote work?

Devore, the founding managing director of the Founders Co-op venture fund, joins us on this bonus episode of the GeekWire Podcast to share his thoughts about what happened, and his optimism about what’s next. “I think Seattle is setting itself up for a great moment in its entrepreneurial journey,” he says.

Listen below, or subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen. Continue reading for highlights, edited for clarity and length.

The ingredients for startup success: The original promise of both Techstars and Y Combinator was such a powerful new idea at the time. It used to be that to start a startup, you needed to raise millions of dollars in venture capital and have a big team. And because of the collapse in the cost of starting startups, because of open source software and cloud infrastructure, there was this moment when all of a sudden people realized, “Hey, it takes a lot less money to start things.”

The limiting reagent in startup success is just extraordinarily talented people. Talented, ambitious people who want to make a difference in the world. And if you can get those people together, and provide them just a little bit of resource, but lots of support and mentorship, they can create magical things.

The ‘give first’ approach: Techstars created this ethos they called “give first,” which is like, “Hey, this isn’t about the money at the outset.” Sure, we’re all investors and we’re in this to try to make things work. There was a real sense of collaboration and community support of ambitious young founders who wanted to make a difference in the world, and the community coming together to help them.

I admit, I’m sort of nostalgic for that moment. That moment came and went with the zero interest rate bubble and the boom and collapse that’s come since. It’s taken 10 years for it all to play out. But there was a golden moment in startup formation that was really a joyful time to be a part of the work.

The broader evolution of Techstars: Our investment firm is called Founders Co-op for a reason. We as a partnership believe deeply in the unreasonable commitment of founders to make a difference in the world. That is really the animated spirit of creating companies. And that’s not typically done in partnership with corporates. It’s saying, “Look, there’s got to be a new way of doing things.”

I think it’s very hard to hold the two ideas at the same time, which is, I’m going to serve the needs of corporates that pay my bills and I’m going to support the needs of founders that want to just pursue the opportunity that they’re pursuing. Trying to stitch those two communities of interest together is essentially a recipe for failure. And that’s the choice that Techstars unfortunately made.

Addressing the reaction to his post: I didn’t intend to attack Techstars. … I have been mourning the death of Techstars for years. I didn’t feel at liberty to say anything because I wanted to be supportive of the program in Seattle and the leadership in Seattle that was trying to carry forward, I think with good intent, all the ideas that we started it with, even as the organization drifted away from its principles.

But not only when Techstars canceled the Seattle program but when they sort of threw the Seattle market under the bus and said, “Oh, well, Seattle’s not important enough for us anymore,” I was like, “Come on you guys, that’s not being honest with us or the rest of the world about what’s really happening. And if you’re not going to be honest, I’m going to do it for you. Not in a mean-spirited way, but in a genuine spirit of honest feedback about where I think you made mistakes.” But again, it clearly made them mad, and I’m sorry I made them mad, but I didn’t feel like I needed to sit on my feelings any longer.

What if someone were starting an accelerator from scratch today? You have to examine everything from first principles, but I think a lot of the conditions that necessitated the creation of a Y Combinator or a Techstars are still true. Particularly with AI, it’s extraordinary how much product a small, lean, lightly funded team can create because of not just open source and cloud infrastructure, which are the unlocks of the first wave, but now you’ve got AI assistance for coding so that even small teams can generate a lot more product.

Where can Seattle’s startup community go from here? At the pre-seed and seed stage market, there’s plenty of wealth in the Northwest to do the work from a capital standpoint. And the same thing is true from a community sense. We don’t need a Techstars or a Y Combinator. There are extraordinary people here, founders, executives, leaders who’ve lived through the startup journey from inception through high growth and success that have a spirit of community and giving back and essentially mentoring other people.

We don’t need to look outside Seattle. We just need to create a framework and a fabric that allows those people to engage in the work in a high quality, high performing way that treats their time and know-how with respect, and also gives them a curated set of companies so that lots of people want to be entrepreneurs.

I think we’ll find, over the next weeks and months, that the community comes together to create another local homegrown version that’s not Techstars in any way, it’s its own thing, but it takes that same spirit of extraordinary founders, high quality curation, and a fabric or a framework that allows people to engage and support those founders in the right way for their stage of performance.

Does an accelerator need a physical location nowadays? Organizations are human constructs. They don’t float in the ether. It takes human leadership, human agency, human collaboration and trust and all those things to get an idea to be instantiated as an organization that builds and sells and delivers for customers. … I think that starting companies begins with humans who look each other in the eye, and that tends to be in the same physical space.

Insights from the Techstars class of 2011, and what its success says about the current economic backdrop: Those founders were starting to think about starting something right at the depths of the global financial crisis. And so the kind of people who are willing to take that kind of risk, against the insurmountable odds of being an entrepreneur, tend to be really gritty, determined founders.

One of the things that has been dispiriting about the last decade or so is, with zero interest rate policies and easy money as a way to get us through the financial crisis, and then through COVID, it started to look easy. Being an entrepreneur started to look easy, and raising money seemed easy and there were all these great stories about unicorns and it tends to bring out less committed, more mercenary and less missionary founders.

What’s next? One of the things I’m frankly excited about the current economic cycle is, a lot of the air has come out of that balloon and I think we’re actually poised for a really exciting renaissance or a moment of excellence like we saw back then not only in Seattle, but in the global ecosystem. These resets are hard and uncomfortable. People lose their jobs. I don’t want to make light of any of that, but I really think it was less about the magic of Techstars Seattle than it was the business cycle and the kind of founders that are willing to take that kind of risk when things are hard.

Founders are people of extraordinary ability and commitment, and much of the good in the world happens from people who are willing to buck the trend. It’s joyful work to support them. I think Seattle is setting itself up for a great moment in its entrepreneurial journey. … I’m feeling very excited for our fund and for Seattle and for all of us who do what we do to support entrepreneurs in Seattle. I think it’s going to be a really exciting couple of years ahead of us.

Related content

• David Cohen: Techstars is evolving and growing
• Techstars: Techstars 2.0: Supercharging Founder Success
• Chris DeVore: What went wrong at Techstars
• GeekWire: Techstars Seattle is shutting down as accelerator shifts focus to cities with more VC activity
• Marcelo Calbucci: The most successful accelerator cohort ever: How this Techstars Seattle class produced 3 unicorns

Techstars Seattle helped launch more than 160 startups over the past decade, partnered with the likes of Microsoft and Amazon to spark even more startups, and led to the creation of three companies currently valued at more than $1 billion — making it one of the most successful programs in the Techstars network. 

Founded in 2006 in Boulder, Colo., Techstars provides fledgling startups with early capital, coaching, mentorship, a chance to pitch to investors, and an opportunity to work for three months in a shared space with other entrepreneurs.

Techstars expanded to Seattle in 2010, and for more than a decade, it worked. And then, this week, it ended. TechStars announced that it’s closing its Seattle accelerator as part of a broader restructuring.

So what happened? And what’s next? GeekWire managing editor Taylor Soper joins the GeekWire Podcast this week to address those questions.

Related stories

• Seattle tech leaders lament departure of Techstars but remain bullish on new opportunities
• Techstars CEO responds to former Seattle managing director, tells him to check his facts
• Techstars Seattle is shutting down as accelerator shifts focus to cities with more VC activity

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

This week on the GeekWire Podcast, stick around to the end for a late-breaking addition to the show — a bonus segment about OpenAI’s new text-to-video technology Sora. Joining us is Rahul Sood, CEO and co-founder of Irreverent Labs, a Seattle-area startup that is working on AI to turn images and text into video.

Also on this week’s show:

• We discuss startups bringing employees into the office five days a week.
• We talk about the upcoming release of a juicy Amazon book.
• We preview the GeekWire Awards in advance of deadline for nominations.
• And we talk about a specific approach to chatting with ChatGPT on the go.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

This week on the GeekWire Podcast, we examine the timing of Jeff Bezos’ plan to sell up to 50 million shares of Amazon stock, in the context of his public announcement that he would be moving to Miami; assess Satya Nadella’s track record as Microsoft CEO as he reaches 10 years in the role; and test an example of how text-to-speech apps could change the way we engage with printed books — or maybe not, based on the way one of us reacts.

After we finished recording, a regulatory filing Friday showed that Bezos sold $2 billion worth of Amazon shares under the plan, his first stock sale since 2021.

Related stories and links

• Microsoft’s resurgence: Reflecting on Satya Nadella’s leadership, a decade after he became CEO
• Filing: Jeff Bezos adopted plan to sell up to 50M Amazon shares six days after announcing move
• Show HN: ReadToMe (iOS) turns paper books into audio (readtome-app.com)

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Audio editing by Curt Milton.

First up this week on the GeekWire Podcast: an inside look at developing software for the Apple Vision Pro spatial computing headset.

We got a chance this week to talk with a longtime software developer for Apple platforms, Ken Case, co-founder and CEO of Seattle-based Omni Group, which makes productivity apps for Mac, iPhone, and iPad, and just released a version of its OmniPlan project management software for the newly launched Apple Vision Pro.

• Related story: Apple Vision Pro: Why one longtime software shop is jumping head-first into spatial computing

Then, we dive into AI, politics, and a new attempt to defuse deepfakes.

A few weeks ago on the show, Oren Etzioni, a University of Washington computer science professor and longtime artificial intelligence specialist, hinted at a secret project in the works. This week, he unveiled nonprofit, nonpartisan technology organization, TrueMedia.org, that is developing an AI-powered tool to detect AI-generated deepfake videos, photos, and audio, aiming to combat political disinformation in the leadup to the 2024 elections.

We jumped back on the line with Etzioni to get the details on the new initiatives, and discuss the pros and cons of the rapid development of new generative AI tools for democracy and society.

• Related story: New nonpartisan AI nonprofit TrueMedia, led by Oren Etzioni, is making a political deepfake detector

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Audio editing by Curt Milton.

Over the holidays, my daughter and I experimented with creating custom nametag stickers for the family members we were traveling with using a cheap portable black-and-white thermal printer and Microsoft’s Bing AI image generator.

When I posted about this on LinkedIn, one of the people who commented was Kevin Leneway, a principal software engineer at the Pioneer Square Labs startup incubator in Seattle, who showed how he used DALL-E-3 (via ChatGPT) to create vivid, colorful illustrated letters to help his daughter learn to read.

Here’s the image he shared:

It was such a interesting example of what AI can do — the type of thing that would take a human artist hours to create, and I appreciated that Kevin was looking for ways to apply AI beyond work, to his personal and family life.

Leneway is a longtime member of the Seattle tech community, the long-ago author of a great blog called A Startup A Day, and a past GeekWire Podcast guest. After our LinkedIn exchange, we ended up catching up via email, with me asking for his insights on various AI technologies for projects we’ve been exploring at GeekWire.

As a software engineer, former Microsoft developer evangelist, and past startup co-founder, his job at Pioneer Square Labs involves working with entrepreneurs to build initial versions of their products, to help get startups off the ground.

Of course, these days, that means he’s working heavily with AI tools.

This has given him an “AI-first mindset,” as he describes it.

“Anything I’m doing, whether it’s a problem I’m trying to solve at work, or even something at home, my first little thought in the back of my head is, huh, I wonder if ChatGPT could help me out with this,” he explains. “Sometimes it works. Sometimes it doesn’t. But I find just having that mindset has unlocked a lot of really weird, and interesting, and creative uses of AI that I would have never thought to do.”

In fact, he has been working on an AI tool of his own, a coding assistant called JACoB, which he introduces in the video below.

The site for JACoB was just launched this week. Leneway and his PSL colleagues are looking for developers to try it and provide feedback. You can sign up via the site.

So that’s how I ended up inviting him back on the show for this wide-ranging conversation about AI from the perspective of someone who is pushing the tools to the edges of their capabilities. We discuss how he’s applying AI to startup ideas and the creative process, how he thinks about AI in his own life and work, big-picture questions about the impact of AI on work and society, and where AI is headed next.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

This week on the GeekWire Podcast, we consider Microsoft’s latest AI product, the $20/month Copilot Pro subscription service, compare it to the likes of ChatGPT Plus, Claude Pro and Otter.ai — and realize in the process that one of us is already paying way too much for AI assistants on a monthly basis.

Plus, our colleague Taylor Soper’s deep dive into the downfall of Zulily bolsters a not-so-far-fetched hypothesis about the motives of the private equity firm that acquired and ultimately shut down the online retailer.

And finally, we compare, contrast, and appreciate the approaches taken by Costco and Amazon to authenticate customers at store entrances and check-out.

Related Stories:

• Copilot Pro vs. ChatGPT Plus: Microsoft’s new paid service offers alternative to OpenAI subscription
• Zulily’s downfall: How the high-flying online retailer soared, sank, and shut down
• Zulily sues Amazon, alleging that price-fixing and supplier coercion sank its attempts to compete
• Scanning against scammers: Costco testing membership card readers at store entrances

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Liz Hamren has been CEO of Amazon’s Ring business for almost a year, as the successor in the role to Ring founder Jamie Siminoff, but it’s not the first time she’s been involved in home cameras. Earlier in her career, she led marketing at Dropcam, where she expanded its user base before its acquisition by Google.

“I honestly have some unfinished business from that time, because we sold early to Google, and there was still so much we wanted to do,” she said in a recent interview with GeekWire. “So I feel like I get to come full circle 10 years later, being able to work in a similar space on home monitoring products.”

Hamren’s position at Amazon builds on a career that spans both technology and marketing. An engineer by training, she has held leadership roles at companies including Microsoft, Facebook’s Oculus, and Discord, contributing to a variety of tech product launches such as Xbox consoles and Meta VR headsets.

At Amazon, in addition to leading the Ring business, Hamren oversees Blink cameras, the Amazon Key service and the Amazon Sidewalk neighborhood wireless network. While the market for in-home cameras has matured over the past 10 years, it continues to be reshaped by new technologies, such as AI and drones.

“There’s so much opportunity for growth,” Hamren said. “With improvements in computer vision and AI, there’s still so much more we can do to provide you with really great information about what’s happening.”

At times over the years, this push to give users visibility into their homes has come with downsides, as evidenced by a $5.8 million settlement between Amazon and the FTC in May over allegations that the company made misleading claims about its privacy and security practices, and failed to protect consumers’ video data from unauthorized access. Amazon disputed the claims even as it agreed to settle the case to put the issue behind it.

Hamren said this intersection of security, privacy, and features is a major focus for the Ring team.

“I think you have to recognize that there is a tension there,” she acknowledged. She cited the company’s efforts to provide privacy to users, give them control over their personal data, ensure accounts and systems are secure, and offer tools to protect the privacy of their neighbors and others around them.

On this episode of the GeekWire Podcast, we talk with Hamren about Ring’s direction under her leadership, the career path that brought her to this role, and how advances such as AI and drones are reshaping this part of the tech industry. She also drops some hints about new Ring features in the works.

Listen above, or subscribe to GeekWire in Apple Podcasts, Google Podcasts, Spotify or wherever you listen. Continue reading for additional highlights, edited for clarity and length.

The current state of the Ring business: We still have an incredibly healthy doorbell business. Our camera business is our fastest-growing, it’s growing at about 30% a year, and the business still continues to be quite healthy. And then we continue to grow [Ring’s home alarm business] as well. And we’ll continue to innovate with things like our glass-break sensor, so that you can really know what’s going on at home when you’re not there.

AI-driven Ring features: We’ve been doing computer vision for a long time, and anyone that’s been in this industry has been. And so it’s actually exciting that now there’s been so much innovation, and also so much awareness, and then just a lot more acceleration of underlying technology that we can leverage. .

• We’ve been continuing to improve what we’re doing with things like person detection, things like package detection. People want to know specifically what is happening and be notified only of those things. …
• Then there’s things like vehicles, which a lot of people have been asking for. So that’s maybe one hint, which is, we’re working on being able to detect vehicles and notify you of that. …
• Video search is coming. That is something that we are working on. … People have a whole host of saved clips. How do you find the thing that you’re most interested in?

Security and privacy: It’s a privilege for us to build these products for customers, and for them to invite us into their homes, and we take that privilege very seriously. And we want to ensure that we’re building tools, we’re being incredibly transparent, and we continue to evolve with new features that really address the needs of our customers, keeping their data secure, making sure they have the ability to customize their cameras and understand how and where their videos are being used.

Status of the Ring Always Home Cam flying security drone: We’re continuing to work on it. We will provide updates as they come. But it is definitely still a project that we are very excited about, that is very ambitious. So stay tuned.

Status of Amazon Sidewalk neighborhood wireless network: Back in the spring, we allowed third parties to use our SDK and order dev kits to start using Sidewalk. We’ve had incredible engagement with third parties and are working closely with a few to build out Sidewalk experiences. So the network is alive and well … and we’re also using it internally for testing different kinds of sensors that makes sense to be on the Sidewalk network.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Audio editing by Curt Milton.

Startup investor Nancy Wang sees the potential for generative artificial intelligence to change the balance of power in fields such as cybersecurity, which is one of her areas of focus, leveraging her background as a former Amazon Web Services general manager and Google lead product manager.

But she has also seen first-hand the ways that AI can unlock opportunities for underrepresented groups in tech, as the founder and board chair of the non-profit Advancing Women in Tech (AWIT).

Collaborating with the U.S. State Department on its Academy for Women Entrepreneurs (AWE) program, Advancing Women in Tech ran a series of classes that culminated in a workshop last fall at an AWE event in Taipei.

Participating in the program were about 75 women entrepreneurs, who did not come from engineering backgrounds but were able to create e-commerce storefronts for their businesses with help from generative AI.

“It lowers the barrier to entry for people, especially underrepresented groups … to be able to create their own livelihoods and be entrepreneurs,” Wang said, describing it as an example of AI’s potential to level the playing field.

That’s one of the takeaways from this episode of the GeekWire Podcast, featuring a conversation with Nancy Wang. A technology product and engineering executive, advisor, and investor, she is a venture partner at Felicis Ventures, where she invests in early-stage startups in cybersecurity, enterprise infrastructure, and business-to-business software as a service. She’s also a contributor to Forbes.

Listen below, or subscribe to GeekWire in Apple Podcasts, Google Podcasts, Spotify or wherever you listen. Continue reading for additional takeaways.

More highlights from Nancy Wang’s comments, edited for clarity and length:

The “virtuous data cycle” in AI: A platform that thrives off user data will only get more valuable the more users and the more data it collects. … If you look at companies that have been born before the generative AI age, they’re now also training their own foundation models, maybe using 10-20 years of, let’s say, network intrusion data, versus a net-new company that will predict network attacks using generative AI. Who’s going to have that deeper moat?

And that represents a lot of the questions or debates that we as investors will have, when we think about the viability of these new companies.

The impact of online learning: The University of Pennsylvania is one of the few universities that has a full Master of Computer and Information Technology degree program on Coursera. The intent behind that is much very much similar to why we, as AWIT, put all of our content on Coursera, which is, we wanted to reach underrepresented students.

If you look at the enrollment rates for the online masters program, it’s nearly 40% to 45% women enrollment, which is unheard of in-person, on campus. So that begs the question: What will we see in terms of the underserved population’s economic abilities if we’re able to provide them these opportunities to flexibly uplevel themselves, to gain more skills, and to then go into more enriching careers?

What drives her work with AWIT: As one of the few women engineering graduates from undergrad, and then going into traditionally very male-dominated industries, such as infrastructure software, data protection, or security … sometimes, I did have to check myself when I was the only woman in a room, and that felt very normal to me. I had to check myself: Wait, this is not normal. This is not what should happen. …

I’m a natural problem-solver. That’s just who I am. I’ve always taken things apart, fixed things, put things back together, probably broken things in the process, as well. I see this as a problem that we need to solve as society. Because if we truly want to advance, and especially using AI products as an example, if we want that to be representative of all of the users, not just people who happen to be building the models, or who happen to be feeding the data, then we do need to bring up some of these underrepresented groups. …

Being the only woman in the room is not a norm that is going to exist much longer.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

“The biggest surprise is that AI used to be narrow. … And when you went from one application to the other, you had to redo it all from scratch. What’s really surprised us is the emergence of general tools. … It’s almost the love child of a search engine, like Google, which you can talk to about anything, with the somewhat limited but very real intelligence of AI systems.”

That’s computer scientist and entrepreneur Oren Etzioni, our guest this week on the GeekWire Podcast, reflecting on the past year in AI, and looking ahead to what’s next.

Etzioni, an AI leader for many decades, is professor emeritus at the University of Washington, Allen Institute for AI (AI2) board member, AI2 Incubator technical director, and Madrona Venture Group venture partner.

Among other roles, he served on the National AI Research Resource Task Force, which advised the White House on policy issues, and he’s working on a project to combat AI-related misinformation in the upcoming elections.

In the first segment of the show, my colleague John Cook and I discuss the big AI news of the week: The New York Times Co.’s lawsuit against Microsoft and OpenAI over their use of its articles in GPT-4 and other AI models.

• Related coverage on GeekWire: NYT v. GPT: Microsoft finds itself on the other side of an industry-defining copyright dispute
• Francesco Marconi in the WSJ: AI and Journalism Need Each Other

Listen below, or subscribe to GeekWire in Apple Podcasts, Google Podcasts, Spotify or wherever you listen, and keep reading for key takeaways.

Highlights from Oren Etzioni’s comments, edited for clarity and length:

The evolution of AI in the past year: “It’s really been a wake-up call. … I like to say our overnight success has been decades in the making. And a lot of us have been aware of the potential of AI for a while. I don’t think any of us anticipated just how quickly and how dramatically, it would come in the shape of ChatGPT, and so on. But we all knew that it was coming. Now, it turns out that the rest of the world, literally, is catching up. That includes the politicians, that includes the kids, that includes the teachers. It’s now changing every aspect of society.”

The role of AI in our work and lives: “Copilot might not even be the right phrase. Maybe an assistant. And an assistant is often only as good as the tasks that you’re able to give it. We cannot delegate some things, we delegate badly, and they’re ill-specified. … I think we have here the potential of finding the drudgery, finding the things you don’t like to do, and having AI give you a big help with those.”

The “toothbrush test” for AI: “What happens next in 2024, is what somebody once called the toothbrush test. So how many times a day do you use the technology. For most of us, toothbrush is two to three times. So I think that in 2024, the toothbrush test for AI is going to explode. We’re going to find that we’re using it two, three, 10, 20 times a day. And I’m not even talking about its implicit use where you’re doing speech recognition in your car, or the Google search engine is using it to re-rank things. I’m talking about us interacting with AI systems, with our music, and our art, or in our job. I think it’s going to be easily 10 times a day on average.”

The need for strong open-source models: “The consolidation of power in AI is a huge risk. And we’ve seen some of that with the top corporations. The countervailing forces are, number one, open source models. A great analogy here is what we’ve seen in operating systems. We had Windows, which billions and billions and dollars went into. But we also had Linux, which the open-source movement championed. So I hope that we will have a Linux of language models, a Linux of AI. And I also think that the government has a role to play in making resources available.

The risks of AI-fueled misinformation: “We’ve already seen it in previous elections, but it’s gotten cheaper, easier to do with generative AI, and I am terrified of its effect on the November election … on the primaries, on the election itself, the potential for distrust, and so on. And I am determined to do something about it to help figure out how generative AI doesn’t become the Achilles heel of democracy.”

What it will take to combat AI misinformation: “I think we need strong regulations. I think we need education. People need to understand how to critically evaluate what they’re hearing, particularly over social media. … In addition, we do need watermarking, authentication, provenance, so we know where things come from. And in addition to all that, we need the ability to detect. So when I see a video, when I hear audio, I have to be able to ask, ‘Was this altered? Was this manipulated? Was this automatically generated by AI?’ With those pieces, I think we have hope to have a robust system. Without any of them, I think we are seeing some major risks.”

The prospects for AI startups: “Some people have the perception that right now, it’s hard to launch a successful startup because of the huge amount of compute power required for these massive models. I think nothing could be further from the truth. We’re at a moment of disruption, and disruption creates a lot of opportunities.”

Advice for aspiring computer scientists: “One, study the fundamentals. Math, statistics, the basic ideas of computer science, those have not changed. Those are the building blocks on which we’re building the latest technology. …

“The second thing I want to say is, follow your passion. So often, people are worried or trying to game the future. Well, I should study this because I could get that job and I should do this. You’re young, the world is changing quickly. Follow your passion, enjoy the educational process, enjoy learning what what you need to do, and these things will take care of themselves.”

[Editor’s Note: We’re excited this week to welcome a guest host to the GeekWire Podcast, Ross Reynolds, whose voice is well-known in the Seattle region from his 34 years at KUOW, the public radio station from which he retired in 2021. He’s joined by a special guest for a look behind the scenes at work and life inside Amazon.]

Kristi Coulter’s latest book is Exit Interview: The Life and Death of My Ambitious Career. It’s a memoir about what she learned in her 12 years at Amazon about work, gender bias, and herself. Exit Interview is Coulter’s second book. Her first, Nothing Good Can Come From This, is a collection of essays about quitting drinking. As you’ll learn from this podcast conversation, it intersects with her Amazon career. 

Listen below, or subscribe to GeekWire in Apple Podcasts, Google Podcasts, Spotify or wherever you listen. Continue reading for excerpts from the conversation, edited for length and clarity.

Ross Reynolds: I really enjoyed the book tremendously. I realized I didn’t know anyone from Amazon at the high level that you were, so I’ve never been able to have this conversation even though I’ve lived in Seattle for 30 years.

Kristi Coulter: Right. People don’t talk, either.

Ross Reynolds: OK. It’s not my fault.

Kristi Coulter: No, it’s certainly not entirely your fault, but people from Amazon have been very closed-mouthed about the place.

Ross Reynolds: Why?

Kristi Coulter: I think part of it is just general discretion. You don’t want to spill, tell how the algorithms work or something, but also I think there’s a certain amount of fear. It’s actually been interesting to me in the past three or four years to see things start to leak out of Amazon pretty routinely. I don’t remember ever hearing about a document leaking to the press in my 12 years there. It probably happened at some point, but now it seems to happen routinely, video footage, documents, all kinds of stuff. ….

Ross Reynolds: The impact of Amazon is what drew you to want to work there, but you had a career before that. What got you to Amazon?

Kristi Coulter: I had a really nifty job at an early internet company called the All Music Guide, which was basically a database of every record on Earth. … I was living in Ann Arbor, Michigan. I had run out of runway in my career. It was just too small a place for me to go any further. And I was really tired of the winters. Weather has often been a motivating force for me. And I thought, “I’m going to have to leave here to go further in my career” because Ann Arbor, it’s a university town. There’s just not the kind of stuff I was looking for.

And so I applied to Amazon on a whim because my husband has family out here, and I thought, “Well, I’ll get my resume polished. This won’t go anywhere.” They called me within two hours. I was here interviewing a few days later, less than a week later, and I recognized it as my chance to make a big change in my life. I was 35 or about to turn 36. Felt like I was stuck and life might be over, which seems hilarious now to think that at 35, but I was like, “This is my chance. Let’s do it.” …

Ross Reynolds: What were you hired to do at Amazon, and what were your first impressions when you came here [in 2006] and walked in the door for those first days?

Kristi Coulter: I was hired to run merchandising, so basically onsite marketing for the books and media stores, so for five different storefronts, and especially hired to make it a better job. They had an incredibly outdated tool set. I mean, the tools were almost non-functional. They broke all the time. People were really miserable. They had been hired as editors, and the job had then shifted to something where their opinions and their knowledge didn’t matter. It was more about just scheduling content. So that was my job. And when I walked in, my first impression was just the chaos. I expected Amazon to be a well-oiled machine, and it was just so chaotic.

Ross Reynolds: What were the signatures of the chaos?

Kristi Coulter: Just people barely able to take five seconds to say hello to me. Really, no ramp up. I interviewed someone the day I arrived. They put me on an interview loop.

Ross Reynolds: To hire someone else?

Kristi Coulter: To hire someone else. I was like, “Don’t do that.” And I was like, “Okay, I’ll go along with it.” My first boss, who was a very nice guy, had basically no time for me. I mean, I probably had, I don’t know, five or six one-on-ones with him and my first six months. So it was just like, “Just figure it out.” And just basically, they give you a list of names of people to talk to and you just do it. So I was on shuttle buses, shuttling all over downtown Seattle, trying to basically put together a jigsaw puzzle.

Ross Reynolds: But you were successful. I mean, you were there for a dozen years. You moved up the ranks.

Kristi Coulter: Well, technically I never moved up. I was never promoted, which became a big sticking point for me. I came in as a level seven. Amazon has twelve levels or something, and I stayed a level seven, but I did get bigger and bigger jobs. And I think it happened first because it was a small enough company that people could see, including people very high up, that I came in and just started to get it together. They could dump me in there, and I was like, “Well, I’ll figure out how to survive here,” because failing was not something I considered an option. And I think I just developed that reputation as someone who could just figure stuff out and delivered some results and people would kind of tap me on the shoulder for various jobs. It also really helped that I came from a liberal arts background.

Ross Reynolds: Really? Why is that?

Kristi Coulter: Because Amazon needed someone who understood language and writing and voice. And they were never really going to hire for that because they, certainly at the time, didn’t want to believe they needed it, but they did. And once I was there, it became apparent that I could help. I could help the site to have a voice again. I could help the merchandisers to actually write better. And so there were a few very smart executives there who were like, “Well, while she’s here, let’s use this.”

Ross Reynolds: So what was the most difficult thing about working at Amazon? What was the biggest challenge for you? And was it the same challenge for everyone there, do you think, or was it particular to you?

Kristi Coulter: Yeah, it’s just a really punitive environment. It’s a culture where there’s not really much of a thank you. You are basically always focused on what you could have done better to the exclusion of what went well. The culture kind of runs on fear. Everyone I knew at Amazon was some level of afraid, from paralyzingly afraid to just a little bit low-level, I’ve learned to live with this afraid, but it’s brutal in that way.

Ross Reynolds: Is that a good thing in some ways, to be a little bit on edge?

Kristi Coulter: I think to be a little bit on edge is fine. And feel like you’re taking risks and things. … In one of the articles on Amazon, someone said, “It’s where overachievers go to feel bad about themselves,” which I was like, yes, indeed. And I think that it gets to a point where you’re just grinding people down. And I don’t actually know that people do their best work when they’re feeling really afraid. You know what I mean? You’re just thinking about survival then.

Ross Reynolds: Well, you write about stacked ranking, and that’s something that would make you really afraid. For people not familiar with it, they basically rank everybody in the department, and the bottom 10% get fired. That’s end of story. I understand Amazon has denied doing this. But in your book, you say they totally did this, right?

Kristi Coulter: Oh, yeah. A couple of years ago, I was reading an article and they denied ever having done stack ranking, and I was just stunned because I participated in that exercise three or four times in my career there. And yeah, you get in and you rank everyone, and that bottom 10%, they’re not fired right then, but they get a very strong message, as we would put it back then, that they needed to improve. And some of those people really did. There were people where you were like, oh yeah, this person’s probably got to go, but there were also people who ended up in that 10% just because they were solid, but more limited in what they could go on to do. And at some point, throwing those people out is just really self-defeating because not everybody needs to be the quarterback. Not everybody needs to be a rock star. You also need people who are not trying to take over the world, but who are really good at their jobs.

Ross Reynolds: You talk about lack of appreciation, but you also talk about some incidents where people said things to you in meetings that were devastating to you. Could you talk a little bit about those experiences?

Kristi Coulter: I think the worst experience I had in a meeting was, I had sold a pretty ambitious plan to transform the merchandising role to the VPs and Senior VP Jeff Wilke in North America. They were very enthusiastic about it. And I went to the International VP [a different executive] to get his buy-in. And he read the doc — the famous thing where people sit in a room and read the doc — really fast. He read it too fast. And I thought, ‘Ooh, this is either really good or really bad.’ And I said, “So, any feedback?” And he said, “Yes, it’s stupid.” Just full stop. And I was like, “OK … can you tell me something more specific about that?” And [he said], “It’s just stupid.”

Ross Reynolds: Where do you go with that?

Kristi Coulter: Yeah. What do you do with that? One of an executive’s biggest roles is to teach. … That’s not how you teach someone. And eventually, he actually denied that the other VPs could have possibly read it. He basically said, Jeff Wilke, junior only to Jeff Bezos, “Oh, he didn’t read it. He just said he did. Because if he’d read it, he would think it was stupid, too.” …

He said this whole room full of alpha males, who are not exactly used to coddling anyone, they all must have been lying. And then he said I was stupid. … I was like, do I throw down, or do I cry? What do I do? And I just tried to just get out of the room without crying, basically. Just very difficult. But yeah, that was probably the worst. It was a direct attack on me. And even worse, it was this grown man with immense power and knowledge just throwing a tantrum.

Ross Reynolds: I’m a big fan of this podcast called Battle Tactics For Your Sexist Workplace. And as I was reading your book, I was thinking about that. Would that executive have said that to a man?

Kristi Coulter: Probably, honestly. Maybe worse. Most people I worked with at Amazon were, at least on the surface, respectful, but there were a few folks … It’s funny, I had a conversation with a friend, an ex-Amazon friend yesterday who said this same guy once yelled at her on a conference call for two hours. I haven’t heard the same stories about him treating men that way, but men got treated pretty badly at Amazon, too.

Ross Reynolds: Was that ever an issue where you thought, I need to raise this gender bias issue?

Kristi Coulter: Who would you have told? … Part of it is that Amazon is so intent, or the system is so intent on having you blame yourself for all of your problems, that I thought, oh, it’s just me. If I were a different kind of woman, I could handle this better. I’d rise above. Also, as a Gen X-er, that’s how I was raised. Nothing can stop you but yourself, which is not really true, it turns out. But also, there was such a denial of any gender bias at Amazon. People were not even willing to entertain the idea. It would’ve been kind of suicidal to make a big deal out of it.

Ross Reynolds: There’s just been a class action suit against Amazon for gender issues. What’s your comment on that, since your book delves into that quite a bit?

Kristi Coulter: Yeah, it’s been interesting. The suit is about huge pay gaps between women in one part of the organization and a man who was at their level, but making [150%] more a year for the same job. Amazon’s comp, like a lot of tech companies’ comp, is really complicated, no pun intended. When you started really affects your base and the stock. I suspect that it could turn out that there’s a non-sexist explanation for the pure comp issues, not that that makes it right, but the women were classified as marketing project managers, and the man was classified as a research scientist, despite the fact that they all were research scientists. And I think that’s really interesting. That drove a lot of the pay difference. And the most interesting thing about the suit is that there was direct retaliation when the women reported it. They went to HR looking for explanations. And one woman who was up for promotion was told that because she had complained, they didn’t feel she was ready for promotion anymore. That’s the big thing for me.

[Editor’s Note: Read the full text of the lawsuit here. Amazon has denied the allegations in the suit, and said it doesn’t tolerate workplace discrimination.]

Ross Reynolds: Did you experience that? Was your job description different than men’s job descriptions who were giving paid more than you?

Kristi Coulter: It’s always hard to know because you’re not supposed to talk about comp, but I did have a time when a man who reported to me was making $60,000 more than I was, despite being a level lower than I was. And I was like, well, this is odd. And I went to HR and just said, “What’s going on?” And nobody was ever able to give me an explanation. And they were like, “Well, it could be a lot of things.” And they also made me feel a little bit gauche for asking. There’s this sense of, “Well, I mean if you want to keep trying to find out, you can, but … “

Ross Reynolds: Can you be a party to the class action suit even though you don’t work there anymore?

Kristi Coulter: No. It’s just three women right now, but they wanted to cover all women from 2016 to ’20 or something in certain jobs. I don’t think I would be a party to it, but I think it could be a nice forcing function to bring some of this stuff out in the light. There was also some data in it about the man who managed all these women taking women’s names off of documents he was presenting, even though they had written the documents or co-written them. Just some really bad stuff.

Ross Reynolds: If approached by an attorney, would you join a lawsuit like this? Do you think you have the goods?

Kristi Coulter: I would certainly talk to the attorney. I’ve never felt like anything that happened to me at Amazon rose to “call a lawyer” level, but if there was a class action and I could benefit from it, I’d have to take that phone call, right?

Ross Reynolds: You were in meetings with Jeff Bezos. What was that like?

Kristi Coulter: Scary. Jeff, in my experience one-on-one, was really cool. I actually really liked him. He’s funny. He’s very engaged. Almost in every meeting he’d say something that would make me go like, “Oh my God,” something revelatory that I would remember. But it’s frightening because he’s so powerful. To be in the room with somebody who’s the wealthiest man in the world. Or he was second or third at the time, was really, really terrifying to me, and I could never quite get past that. It was very hard to just see him as a person, even though he’s actually quite personable.

Ross Reynolds: Notwithstanding Amazon’s manifest success as a business, from the inside, did you ever think, they could be doing some of these personnel things better? Or does its very success say, “No, they’re obviously doing it right”?

Kristi Coulter: In some ways, I thought the success said they were doing it right. Even at its best, it would not be a place for everyone. You need to be super-comfortable with change and things moving fast and ambiguity, and it’s really exhilarating. I got to work on things I never would have at more sane companies or companies with less money, but yeah, I mean, you would have to look around and think, these people are so good, so brilliant. My coworkers were so talented. What if they weren’t also fried and exhausted and desperate? They could probably be even better, and it just didn’t seem sustainable.

Ross Reynolds: But how do you do both, I guess?

Kristi Coulter: Yeah. I don’t know that you do. I think that it’s interesting seeing the struggles Amazon’s having now. Part of me thinks, have they hit some wall? I got to a point where women locally especially would tell me, “Yeah, I just won’t take their phone call. I won’t interview at Amazon.” The reputation was so bad it was pushing talent away. And that’s anecdotal, of course. I’m sure lots of people would be happy to work there, but I remember thinking, that’s not good when your rep is so bad that talented people just are writing you off because there’s a lot of interesting things about working there.

Ross Reynolds: You write that success at Amazon wasn’t necessarily having these core skills. It was figuring out how to work the Amazon system, how to get the programmers to do what you needed them to do.

Kristi Coulter: Yes. One reason I was able to move around a lot and actually have, really, a few different careers at Amazon was because people would openly say, “Well, I’m not so worried that you don’t know how the publishing business works. You know how Amazon works and you could figure out the publishing business.” And it’s true, it was true. And when I talk to young people now, I always say, “Don’t underestimate your transferable skills. Knowing how to meet people and ask questions and think critically, they’re really valuable.” But yeah, it would take someone else six months to figure out how Amazon works. It took me at least six months, and so we always wanted to hire from within. People would get frightened about hiring from outside because the ramp-up time would always be longer than we thought we could afford, because that’s just how humans work.

Ross Reynolds: Amazon is so massive, and that really came through to me in a section in the book where you read news accounts, these famous accounts, of workers at an Amazon Fulfillment Center being taken out in ambulances because the temperatures were over 100 degrees. And part of your reaction was, “Oh yeah, that’s the company I work for.”

Kristi Coulter: Yeah. That was [reporter] Spencer Soper. He’s here in Seattle now, but he wrote that in Pennsylvania. Yeah. I was like, “Oh, right. We have warehouses.” Because I moved away from the retail business fairly early on. … And it was shocking to realize how quickly something like that could just fall out of my view. But it did and then it did again.

Ross Reynolds: You went there because you were ambitious to do something big. By the time you left, did you accomplish what you set out to do?

Kristi Coulter: I worked on things I never would’ve dreamed of. I think in some ways I did accomplish what I set out to do, but I was the kind of person who got promoted like clockwork my entire life. I was just that kind of girl, I’m going to do the extra credit, I’m going to get promoted, and I never got the big promotion that was dangled in front of me for 12 years. I think I had seven or eight conversations with different bosses that were like, “You’re a year away.” And so by that one external metric of success, I left feeling like a failure. I was like, “Twelve years. These people, no one could manage to get me promoted.” And it honestly still kind of bugs me.

Ross Reynolds: So you internalized that. You thought to yourself, “That must be my fault that I didn’t get the promotion.”

Kristi Coulter: I definitely thought so for a while, and then at some point I was like, “These people don’t have their act together.” Nobody could give me the same story. I had one VP tell me — I had a whole document with what I thought I needed to do to get promoted, and I wanted his feedback and he said, “Just change the world. Just change the world and you’ll get promoted.”

Ross Reynolds: Just change the world.

Kristi Coulter: Yeah. I was like, “Oh, thanks, dude. That’s great.” And it was like he wasn’t even trying to have a serious conversation with me. And so I got to a point where I was like, “These people don’t understand what they’re doing, what they’re talking about.” I stopped blaming myself a bit then.

Ross Reynolds: I’m sure people ask you, “Should I go to work at Amazon?” What do you tell them?

Kristi Coulter: I usually tell them, “It depends.” I don’t tell people, “No. Just run screaming.” But I think you want to be very specific and clear about what you want out of the experience. You want to probably go in with an exit strategy. Maybe if you’re young in your career, you go in and you say, “I’m going to stay for three years and I’m going to do X, Y, and Z, and then I’m going to go.” And you have to remember that it’s going to be rough and that no one is really going to care about you. Your coworkers and your boss, lots of wonderful people work there, but the system of Amazon really doesn’t care if you are a well person or not, and it will absolutely spit you out.

Ross Reynolds: Since you started, Amazon has all these brand new buildings in downtown Seattle. So has the catering improved? Have those elements that were so starkly not there when you arrived, did they change over the years?

Kristi Coulter: A little bit. There are cafeterias now. I was in Columbia Tower with that weird food court, but there’s no gym, the basic corporate amenities. We had bike cages and eventually we had some locker rooms. There weren’t rooms for nursing mothers when I got there. There was a bathroom.

Ross Reynolds: Were there by the time you left?

Kristi Coulter: Yeah, yeah. … But I was at a coffee shop somewhere and I heard a guy running a startup talking to an employee, and he was trying to give her a pep talk, and he was like, “This isn’t Amazon. I’m not going to be able to give you the free meals and the free haircuts and the massages and … but it’s going to be fine.” And I was thinking, wait, he’s got the wrong idea about Amazon. It’s not Google. So the main perk for employees was you could get $100 off the website per year if you went into the HR tool and found this code in January. Nobody would email you to remind you. They wouldn’t send it to you. It was basically, if you remembered you could get that $100 off.

Ross Reynolds: Was that a conversation among employees there? “Boy, they’re kind of cheap!”

Kristi Coulter: Yeah. We had a leadership principle of frugality, which is that you want to be frugal with money, which makes sense. But at some point, someone made a Wikipedia page called “frupidity,” for the merger of “frugal” and “stupid.” There were things like people being expected to fly to India multiple times a year in coach and go straight to the office when they got off the plane. I had a computer that was taking seven minutes to boot up and they didn’t want to give me a new one because it wasn’t quite at the end of its four-year life cycle. Things got really stupid.

Ross Reynolds: Your first book: Nothing Good Can Come From This: Essays About Quitting Drinking. Your second book, Exit Interview, The Life and Death of my Ambitious Career. Are those two things related, the drinking issues and working at Amazon?

Kristi Coulter: Yeah, absolutely. I mean, Amazon did not make me an alcoholic. I was probably destined to be an alcoholic from a very early age, but I certainly was drinking more and more to just cope with the stress of the day. It was the only way I could forget every mistake I had made that day or every perceived mistake, and it really, really ramped up. And I actually ended up doing an A/B test because I quit drinking halfway through my Amazon career, and I wasn’t sure I’d be able to stay, and stay sober, but I did. I stayed another five or six years. It’s doable, but I really had to put up boundaries and develop a spine in a way that I had not had to as a drinker, when I could just go home and drink it away. So I had to change as a person, and not everybody liked Kristi with a backbone.

Ross Reynolds: That shows enormous strength, not only to be able to overcome this difficult workplace, but to overcome alcoholism, which is a disease.

Kristi Coulter: I didn’t want to die. I was either going to die young or I was just going to not — My life was getting smaller and smaller. It was just Amazon and then drinking to get over Amazon, and I had enough, I don’t know, vision to see that there had to be a better way.

Ross Reynolds: So have you gotten any comments from former colleagues at Amazon about Exit Interview?

Kristi Coulter: A flood. A flood of them. I have heard from so many people in Amazon offices around the world, women especially, saying, “Thank you for telling my story.” But a lot of men too, which surprised me, because the book has such a gender inflection. Men saying either, except for the gender part, you were telling my story. … Or, there’s a lot in the book about the sexism as a lot of unconscious bias. It’s not a place where men are just like, “Well, women are dumb.” They don’t know that they’re sexist.

Ross Reynolds: Could you give an example of that? What is something that was, obviously, they’re just clueless?

Kristi Coulter: The leadership at that level. So at my level, it was 20% women and 80% men. Amazon at entry level is about 50/50 split, and then women just vanish. And you get into management, women are gone. For most of the time I was there, there was no woman reporting directly to Jeff on the S Team.

Ross Reynolds: The S team is? For those who don’t know.

Kristi Coulter: Oh, yes. The S team is basically Jeff’s direct reports. It’s the very top level of leadership. But whenever this would come up, men would just be like, “Well, I guess women, they just don’t want these jobs really.” Or, “Women have different priorities.” … I was like, really? It’s 80% male as you go up the ranks? And a lot of it would come around to, well, women have children. But I was like, but I don’t have… In fact, a lot of women I knew at Amazon did not have children because if you wanted to rise, it was much easier not to have children.

But I think that men were falling victim to this thing that men do sometimes and that I see white people fall victim to also, which is thinking it was their personal fault if an environment had structural sexism.

So they weren’t able to step back and be like, “This is weird.” Amazon had no company daycare, and that was something women talked about a lot, and I rarely heard men talk about it, and if men had talked about it, maybe something would’ve changed because they had the power. But it was just like, well, women don’t want these jobs. I heard many times, “Oh, women are too smart to want these jobs.” Like, oh yeah, we’re just passing on the power and the money and the leverage. Sure. We’re just too smart for that.

And so it’s frustrating because when you can’t name a problem, you can’t talk about it at all. So a lot of men wrote to me saying, “Thank you for showing me what I just… absolutely, it was not seeing. I was just convinced that this is just the way that nature made it. Clearly, it’s supposed to be mostly men for a reason,” and that was great to see men be like, “Oh, right, this is not normal.”

I’ve also had women tell me that physical copies of the book are being passed around Amazon from woman to woman, and they’re writing notes and inscriptions to each other, and that sent chills down my spine. It’s like a new little whisper network for women, like a yearbook or something.

Ross Reynolds: Have you got to do any readings here in Seattle and have those people show you those books with the annotations?

Kristi Coulter: I’ve seen a couple photos of one of those. But yeah, I have women come up to me at events in Seattle and within a minute they’re crying. They just walk up and just start sobbing. And it’s heartbreaking. I mean, it’s wonderful because they’re crying because they feel like the book showed them themselves, but I feel like there’s a lot of people in pain out there in tech.

Ross Reynolds: Your book is called Exit Interview. Is this instead of a real exit interview? Did you get a real exit interview at Amazon? Did you get to say any of this?

Kristi Coulter: No, I did not. Amazon, it’s a huge company. A lot of people just get this form to fill out, but some people do get in-person exit interviews. And as someone who’d been there for 12 years, I was in the 98th percentile for tenure and one of few women at my level. I thought, well, someone’s going to want to talk to me. I was leaving on good terms. I got the form instead, and I was kind of stunned, but I thought, okay, I’ll do this. And we had a tech glitch when I was submitting the form, which I had taken two hours to fill out, and it was lost to the ether, and I laughed because it was perfect. I was like, of course, now I’m going to leave, and still, my voice will not be heard. I could be glib and say, I wrote this book because I did not get an exit interview. But there was something to that. I was like, I have things to say about my experience here, and I am going to say them somehow.

Ross Reynolds: Just a personal question, because you worked at Amazon for 12 years, do you never have to work again?

Kristi Coulter: Oh my God, I wish. No.

Ross Reynolds: Not that good?

Kristi Coulter: Yeah. I was very well paid at Amazon. The pay was great, and I was there at a time when the stock was going insane. No. It bought me some ramp. Amazon paid for me to write this book, essentially. But no, I do need to work, sadly. But I like to work, also.

Exit Interview: The Life and Death of My Ambitious Career, by Kristi Coulter, is published by MCD, an imprint of Farrar, Straus & Giroux. It’s available wherever books are sold (yes, including Amazon).

Production assistance on this episode from Curt Milton.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

• Embrace an “outsider’s perspective” to bring fresh ideas to traditional fields.
• Tap deeply into your personal experiences to inform and inspire your work.
• Use data and facts to shape behaviors and outcomes.
• Capitalize on the human desire for recognition and competition.

Those are some of the insights from GeekWire’s profiles of six “Uncommon Thinkers”: Seattle-area inventors, scientists, technologists, and entrepreneurs transforming industries and have a positive impact on the world.

This editorial series, presented in partnership with Greater Seattle Partners, was based on the deliberations of a panel of outside judges who chose the Uncommon Thinkers from nominations submitted by GeekWire readers.

Working on these profiles helped us identify the commonalities among our honorees, including their inspirations, mindsets, and approaches. Continue reading for a link to each profile, plus a summary of key takeaways.

You can also hear conversations with several of our Uncommon Thinkers on this week’s GeekWire Podcast, recorded backstage at the GeekWire Gala in Seattle, where we honored them during a reception earlier this week.

Read the six “Uncommon Thinkers” profiles here:

• UW prof and entrepreneur Shwetak Patel has a rare ‘creative brilliance’
• Elizabeth Hansen uses her job as an anesthesiologist to cut carbon emissions
• Boundless CEO Xiao Wang on his quest to fix the immigration process
• For USAFacts’ Poppy MacDonald, ‘facts are paramount’ for democracy
• Blake Resnick finds a larger purpose in tech, making drones for public safety
• MagniX CTO Riona Armesmith geeks out over electric aviation

And here are some of the key insights we took away from our interviews.

Embrace an “outsider’s perspective” for fresh ideas: Shwetak Patel, a University of Washington computer science professor and Google distinguished scientist, doesn’t have a traditional background in medicine or electrical engineering.

However, he benefited from his role as an outsider in coming up with ways to use electrical wiring as a wireless communications network for smart home devices, and smartphones as medical diagnostic devices.

Growing up in Alabama, he gained a wide range of electrical and mechanical experience fixing air conditioning units, vending machines, and all sorts of other equipment in the motels managed by his family.

That hands-on experience with the trade, without formal training, was “the thing that helped me ask questions that I don’t think any scientist would have asked, given their scientific training in that space,” he said.

Let your personal experiences inform and inspire you to change the world: Patel isn’t the only one of our Uncommon Thinkers whose specific life experience has heavily influenced his work. In fact, this is one of the strongest commonalities among the honorees.

• Boundless CEO Xiao Wang’s personal challenges during his family’s immigration process inspired him to simplify the immigration system, making it more accessible and transparent as CEO of the Seattle startup.
• Dr. Elizabeth Hansen, an attending anesthesiologist at Seattle Children’s Hospital, was motivated by her concern for the future health of children to implement environmental initiatives in healthcare, reducing emissions.
• Poppy MacDonald’s experience as a media executive and belief in the importance of informed citizenship inspired her to lead the team at USAFacts providing insights into government spending and operations.
• Blake Resnick’s decision to start his drone company, Brinc, was significantly influenced by the deadly mass shooting at the Route 91 Harvest music festival in his hometown of Las Vegas.
• Riona Armesmith’s vision for electric aviation drives her to overcome technical barriers and push the boundaries of what is currently possible in aircraft propulsion as CTO at electric aviation company magniX.

Prioritize transparency: Use data and facts to shape behavior and outcomes.

• MacDonald champions the use of comprehensive, factual data to inform citizens and policymakers, enhancing public discourse and decision-making.
• Hansen uses software from AdaptX, the startup founded by her Seattle Children’s Hospital colleague Dr. Dan Low, to track emissions data from anesthetics, quickly showing the impact of specific decisions.

Capitalize on the human desire for recognition and competition. Hansen took that data transparency a step further by displaying and sharing the emissions results for her entire team in an easy-to-read plot so that everyone could see how their emissions compared to their peers, and to their own past results.

“That brought out everyone’s natural competitive spirit,” she explained.

She would highlight the lowest emitters from the prior month in her email out to the group. People would ask her how to get highlighted in the next month’s email, creating an opportunity for Hansen to show them the small adjustments in their anesthesiology practices that could significantly impact their results.

Tackle adjacent problems: Rather than just trying to solve the main problem directly with many solutions, Patel’s approach is to find all the related “adjacent” problems and try to solve those in parallel.

By solving adjacent problems, he explains, you end up with different ways of solving them that can then be brought back together to solve the primary problem. This allows him to tackle problems from new angles.

Look for opportunities to say “yes.” Startups are hard, and you hear “no” a lot, so when someone believes in you and says “yes,” you have to appreciate that moment, said Wang, the Boundless CEO.

Speaking to his startup specifically, he said, when you witness the daily struggles of immigrants trying to create better lives for themselves, it’s hard not to say “yes” whenever there’s an opportunity to help them.

“For me, being able to be default ‘yes’ means that you can quiet all your fears of failure, you can inspire hope in others and you can appreciate just how darn fortunate we are to be able to say that word,” he said.

Find your inspiration to change the world for the better: Ultimately, this is the biggest common thread among the six “Uncommon Thinkers”: their recognition of broader societal problems, their fierce desire to solve those problems, and their ability to find unique solutions in their own expertise and experiences.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Conference attendees in Las Vegas are accustomed to encountering people standing on the street, aggressively trying to hand them flyers for, well, let’s just say for a variety of different services.

So it was a surprise, and a sign of the times, when someone standing on the Las Vegas Strip handed me a flyer for something else entirely during the Amazon Web Services re:Invent cloud conference this week.

“NEED GPU?” it read, in bold letters.

The promotional materials were from a company touting what it described as the “largest stockpile of GPUs available for the best price” — underscoring the soaring demand and short supply of the graphics processing units instrumental for training artificial intelligence models.

It’s a new world, and of course AI was the big theme inside the longrunning AWS re:Invent conference, as well. The company’s announcements included a new AI assistant for work called Amazon Q.

Before heading back to Seattle, I got a chance to speak with Matt Wood, AWS vice president of product, about the new service and the new era of AI for business.

Several of his comments illustrated the rapid evolution of technology this year:

The rise of generative AI this year: “I have not seen this level of energy and enthusiasm and excitement and engagement from customers probably since the very earliest days of AWS when it was really kicking into high gear. We were thrilled but really having to work hard to keep up with customers’ enthusiasm for building. Very, very similar this year, with generative AI.”

AI and regulated industries: “One of the surprises to me over the past 12 months are the regulated industries. These are industries that potentially don’t have the best reputation for being in the vanguard of technology: insurance, financial services, health care, life sciences, all these sorts of things. But in an ironic twist, all of the regulations that they’ve been working through over the past 20-30 years have been around data privacy, and data governance, and data standards, and structured data quality. They’re all the things you need to have in place to successfully apply your own data to generative AI. … For a very small incremental investment, they can start applying that data with generative AI really, really quickly; really, really easily.”

Changing perspectives on data structures: “I was joking with a customer yesterday, they were saying that they have really, really old data. And a lot of it is structured, but they’ve got this notes field, and it’s just natural language text. And we were joking that, even a year ago, you would have said, ‘I’ve got all this structured data, thank goodness.’ But now you’re like, ‘Oh natural language text, thank goodness, that’s the gold, right?’ So it’s totally changed how you think about the data and where the value is.”

One of the notable aspects of Amazon Q is its ability to connect to business data from a variety of applications, including Microsoft 365, Slack, Salesforce, Dropbox, and Amazon S3. Asked about the execution of these features, Wood explained that the connections are made through APIs, but he also said there’s more cooperation among cloud rivals than people might expect.

“We have good relationships with with a lot of folks that we compete with in other areas. We sell a lot of Windows licenses on EC2 as an example, and through Amazon WorkSpaces,” he said. “So we have we have really good working relationships. The integration as it’s done under the hood is done through APIs, but that is not because we have any animosity towards these things.”

Hear more highlights from my conversation with AWS VP Matt Wood on this week’s GeekWire Podcast. Listen above, or subscribe to GeekWire in Apple Podcasts, Google Podcasts, Spotify or wherever you listen.

Postscript: On Friday afternoon, the newsletter Platformer reported that some Amazon employees are raising concerns about accuracy and privacy issues involving Amazon Q, which they say have come up in the internal use. Q has been “experiencing severe hallucinations and leaking confidential data,” according to leaked documents, including “the location of AWS data centers, internal discount programs, and unreleased features,” as Platformer reported.

Amazon downplayed the issues and specifically denied some of the claims.

“Some employees are sharing feedback through internal channels and ticketing systems, which is standard practice at Amazon. No security issue was identified as a result of that feedback,” an Amazon spokesperson said in statement to GeekWire. “We appreciate all of the feedback we’ve already received and will continue to tune Q as it transitions from being a product in preview to being generally available.”

The statement added, “Amazon Q has not leaked confidential information.”

In the final segment of this week’s show, we listen back to AWS CEO Adam Selipsky’s thinly veiled jabs at Microsoft and OpenAI during his re:Invent keynote this week. You may notice that one of those clips is of lower quality. That’s because it’s from my own recorder, sitting in the crowd. The reference was removed from the official Amazon recording after the fact.

Update: Amazon got back to me on this. This portion of the keynote isn’t available for on-demand viewing because they don’t have permission/license to display the CNBC content in the on-demand version. In other words, they weren’t intending to pull any punches by not putting the comment online.

Related Coverage:

• Sphere of influence? Google’s big Vegas ad buy clouds the landscape at Amazon re:Invent
• AWS unveils ‘Amazon Q’ AI assistant, jabs at Microsoft and OpenAI at re:Invent
• Inside the AI chip race: How a pivotal happy hour changed Amazon’s strategy in the cloud
• Amazon puts Fire TV Cube to work, repurposes streaming device for desktop computing

This week on the GeekWire Podcast, we assess the outcome of the drama at OpenAI, considering what the removal and ultimate return of its CEO Sam Altman, and everything that happened in between, means for Microsoft.

While Microsoft CEO Satya Nadella earned kudos for masterfully navigating the mess, the situation also showed how the company was vulnerable to being blindsided due to OpenAI’s convoluted structure.

And while for-profit interests are likely to get more influence on OpenAI’s reformulated board, it’s not yet clear if Microsoft will end up with a seat on that board, leaving the ultimate outcome for Microsoft unclear, for now.

Another big question: will companies think twice about going all-in with OpenAI and Microsoft now? That could create more opportunity for the likes of Amazon, Anthropic, Google, and others in enterprise AI, in particular.

Also this week, we discuss Amazon’s move into car-buying, and the Seattle region’s dwindling flock of unicorns.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

OpenAI CEO Sam Altman was as bullish as ever about artificial intelligence and the ChatGPT maker’s latest advances when he spoke Thursday at the Asia-Pacific Economic Cooperation (APEC) Summit in San Francisco.

“I think this will be the most transformative and beneficial technology humanity has yet invented,” Altman said, adding later, “On a personal note, four times now in the history of OpenAI, the most recent time was just in the last couple of weeks, I’ve gotten to be in the room when we push … the veil of ignorance back and the frontier of discovery forward.”

Getting to do that, he said, “is the professional honor of a lifetime.”

One day later, Altman was out.

The surprise removal of the OpenAI CEO on Friday is the buzz of the tech industry — raising questions about the company’s role in the AI revolution, the ambitions of Altman and his team, the impact of his exit on the rest of the tech industry, OpenAI’s complex corporate structure, and its unusual partnership with Microsoft.

Microsoft has invested billions of dollars in OpenAI and has used its technology to put Microsoft’s own products at the forefront of the AI revolution.

But the Redmond company was blindsided Friday by the OpenAI board’s decision to remove Altman, learning about it shortly before the news was announced, according to a report by Axios, which GeekWire confirmed independently.

Microsoft does not have a seat on the OpenAI board. The unusual arrangement is a quirk that resulted from OpenAI’s origins as a non-profit organization, OpenAI Inc., dedicated to building artificial general intelligence for the advancement of humanity.

In 2019, the non-profit created an affiliated “capped-profit” entity, OpenAI Global LLC, aimed at giving it the resources to pursue the AGI vision without the profit incentives of a traditional company.

Reports overnight indicate that a conflict between Altman’s ambitions for OpenAI and the mission of the nonprofit were at the core of the board’s decision to oust him as CEO and remove OpenAI co-founder and President Greg Brockman as chairman.

Brockman was initially designated to report to Mira Murati, the OpenAI CTO who was named interim CEO by the board, but he then resigned from his executive role later Friday afternoon.

Based on her reporting, journalist Kara Swisher reported that the problem “was a ‘misalignment’ of the profit vs. nonprofit adherents at the company,” noting that OpenAI’s big moves at its first developer day last week were part of the problem.

At the event, on Nov. 6, Altman announced customized and personalized GPTs and plans for a related marketplace, and told reporters that the company might even come out with its own AI device in the future.

In an interview with the Financial Times, published Nov. 12, Altman said OpenAI would need to raise “a lot more” funding from Microsoft and others, over time, to finance the computing power for future AI breakthroughs. In the same interview, Altman also acknowledged that the company is working on GPT-5, the next major version of its large language model.

The board alluded to a misalignment in its announcement of Altman’s removal, saying that a review had concluded he was “not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities.”

“OpenAI was deliberately structured to advance our mission: to ensure that artificial general intelligence benefits all humanity,” the board said in its statement. “The board remains fully committed to serving this mission. We are grateful for Sam’s many contributions to the founding and growth of OpenAI. At the same time, we believe new leadership is necessary as we move forward. The board no longer has confidence in his ability to continue leading OpenAI.”

According to a transcript of an employee town hall Friday afternoon, obtained by The Information, OpenAI board member and co-founder Ilya Sutskever told employees, “This was the board doing its duty to the mission of the nonprofit, which is to make sure that OpenAI builds AGI that benefits all of humanity.”

The Information reports that Altman’s exit was preceded by internal debates over whether the company was taking enough safety precautions in its development and deployment of artificial intelligence.

The reaction from Redmond

OpenAI’s capped-profit company does not have a board of its own, which is why Microsoft does not have a seat on the board.

It did have an indirect presence on the board previously, when LinkedIn co-founder Reid Hoffman was on the OpenAI board. Hoffman, who sold the business social network to Microsoft in 2016, is also a Microsoft board member. However, he stepped down from the OpenAI board in March.

Following the news of Altman’s abrupt departure, Microsoft CEO Satya Nadella issued a statement, offering reassurances to Wall Street and Silicon Valley about the company’s relationship with OpenAI and its ability to capitalize on the AI revolution with products of its own. He did not directly mention Altman.

“We have a long-term agreement with OpenAI with full access to everything we need to deliver on our innovation agenda and an exciting product roadmap; and remain committed to our partnership, and to Mira and the team,” Nadella said.

But the corporate structure isn’t the only part of the relationship that’s complicated.

As part of its strategy under Altman, OpenAI has begun pursuing enterprise customers directly. During a media Q&A at OpenAI’s DevDay on Nov. 6, Altman was asked how Microsoft and OpenAI were balancing their parallel and potentially competing quests to sell the same core AI technology to enterprise customers.

“The answer is, we’re both going to do it. We set up the relationship between two of us so that we’re very happy when they succeed with a sale, and they’re very happy when we succeed with the sale,” Altman said. “I’m a huge believer that incentives are superpowers. If you get everybody’s incentives right, everybody just wants the most shared success possible. And I think we’ve designed that very well.”

Whether that’s how it actually plays out remains to be seen.

Along these lines, questions about OpenAI’s future could work to Microsoft’s advantage, if enterprise customers decide it’s safer to access GPT-4 and other OpenAI technologies through the Azure OpenAI service and other Microsoft offerings, rather than doing business directly with OpenAI through its ChatGPT Enterprise service, for example.

But that’s a silver lining, at best. Microsoft shares fell more than 3% in after-hours trading Friday, underscoring the importance of the OpenAI partnership to the company, and the impact of the uncertainty created by Altman’s ouster.

Dan Ives, an analyst at Wedbush, said in a note to clients that Saturday that there shouldn’t be any impact on Microsoft’s products and its Copilot enterprise strategy in the short term.

“That said, we would expect Microsoft now to exert more control of OpenAI given its importance and expect more disruption from a OpenAI board/company perspective over the coming months,” he added. “This is the time for Microsoft to now essentially take this over strategically speaking.”

Broader industry implications

Microsoft was wise to speak out quickly, said Matt McIlwain, a managing director at Madrona Venture Group in Seattle who focuses on areas including applied artificial intelligence and intelligent applications.

“What they’re saying is, ‘We’re the ones with all the enterprise customers, we’re the ones that are putting out all these copilots … we’re standing in front of our capabilities … and so you can trust us (Microsoft) even if there’s some instability with one of our key partners.’ ” McIlwain said.

Ultimately, he said, the instability at OpenAI underscores the need for a diversity of AI foundation models in the industry, with cloud platforms such as Microsoft Azure, Amazon Web Services, and Google Cloud offering their customers a broad selection of AI technologies.

AWS has been a leader in this regard with its Bedrock service, providing access to a variety of foundation models. In that regard, McIlwain said, the shakeup at OpenAI could provide a new opening for Amazon: “I think their opportunity is to say, ‘Hey, we’ve been about price, convenience and selection.’ That’s the Amazon thing.”

Microsoft has also been moving to expand its offerings further beyond OpenAI, including the announcement this week at its Ignite conference of expanded availability of Meta’s Llama 2 model through a new Azure AI service.

McIlwain said this week’s OpenAI news creates “an opportunity for a bit of a rethinking by some customers as to who they’re going to partner with deeply” in applied artificial intelligence.

“I think the cloud players, and specifically Microsoft and Amazon, are going to do very well in this world,” McIlwain said. The key, he said, is to make it clear that “they are providing a lot of choice and options to customers — because you just truly never know quite what’s going to happen.”

Matt McIlwain discusses the OpenAI news on this week’s episode of the GeekWire Podcast. Listen above, or subscribe to GeekWire in Apple Podcasts, Google Podcasts, Spotify or wherever you listen.

Fei-Fei Li‘s new book is the story of her journey from China to the U.S., from small business to Big Tech, and from academic research to corporate life, and back again.

But more than that, it’s the story of artificial intelligence, as told through her experience as one of the people summoning this new day and standing there awestruck, excited, and concerned about what it will mean for humanity.

Dr. Li joins us on the GeekWire Podcast to discuss her book, The Worlds I See: Curiosity, Exploration, and Discovery at the Dawn of AI, from Moment of Lift Books, an imprint from Melinda French Gates and Flatiron Books.

“Every time a technology is powerful enough to become products, to become applications, to become horizontally impactful to so many people … it is messy. It is very messy,” she says. “And this is where I think we should take a breath and recognize this messiness, embrace this messiness.”

Ultimately, she says, it’s most important to ensure that technology operates in service to people. On balance, she says the rise of generative artificial intelligence seems to have moved us further toward that goal by spurring the conversation about the need for human-centered AI.

“There’s noise, there’s a lot of hyperbole, and it’s a necessary phase, but I’m hopeful that the human-centeredness is much more front and center now,” she says.

Known for her foundational contributions to AI and computer vision, Dr. Li is the inventor of ImageNet, a large-scale dataset of images that enabled rapid advances in deep learning for visual recognition. She is a professor of computer science at Stanford University and a co-director of the Stanford Institute for Human-Centered Artificial Intelligence, who worked as Google Cloud’s chief scientist for AI/ML during a 2017-2018 sabbatical.

I’ll be speaking further with Dr. Li about her book, The Worlds I See, on Monday evening Nov. 13 at Town Hall in Seattle. See this site for details and tickets.

Listen above, or subscribe to GeekWire in any podcast app. Continue reading for edited highlights.

Perseverance and scientific discovery: “Just like any other thing that’s hard in life, you have to grind. It’s teamwork. It’s both creativity and perseverance. It’s grit and passion. And this is why I deliberately choose to give the kind of details of the journey as this book did.

“And my message to anyone, especially young people who are going through this phase of their journey is, don’t give up. Be resourceful, be creative. Be willing to know when your tool is wrong, or your first idea, your second idea, your 15th idea is wrong. But don’t give up that North Star. Don’t give up that dream, and you persevere if you believe in it and are passionate and work for it.”

Private companies vs. public institutions in AI: “This is actually the challenge of our time. I think this is a critical question. I cannot overemphasize enough that we have a terrible imbalance right now. The imbalance is not only on the resource front, it’s also the voice, the megaphone. …

“Policymakers are meeting with business leaders left and right, which is fine, it’s good they meet with them, but they also need to hear from academia and the public sector. AI is a very powerful technology. It can help us to actually discover more critical science, a cure for cancer, fusion, there’s many things, but it also can be used to optimize advertisement placements and revenue. The public wants the former, some companies want the latter, but we definitely want both to be healthy in our society. And right now, we only have the companies using AI, we don’t have enough resources for the public institutions.

“On top of that, we are worried about the catastrophic or existential risks of AI. Who has the resources to open the hood and examine what’s going on? You need trusted public sector partners. You cannot just completely rely on self-reporting by these companies. In order to do that, you need a healthy academia and public sector.

“Last but not least, who’s benchmarking? Who is assessing? Who is evaluating? Not only evaluating on the speed and performance, but also fairness, privacy, hallucination, alignment, all these issues that we are seeing in today’s technology. And again, we need public sector and academia’s thought leadership in this.

“So for all these reasons, I think we are really in a bit of a crisis if we overlook the public sector at this moment in AI’s development.”

The origins of her human-centered approach, in her experience as a small business owner with her parents: “When I was building the technology, especially when I was seeing the link between AI and healthcare and also other industries, it [was] so easy for me to understand people from the other side, because as a small business owner for a dry cleaner shop, everything you’re trained on is to understand your customers and make sure they’re happy.

“And it really made me understand the struggle of someone, as an immigrant as well as the receiving end of a product or service, the customers and users. So when I was working in healthcare with AI or other businesses at Google, it was second nature to me to try to ground technology in human perspective. So that was helpful.

“And besides, I live in Silicon Valley. I can tell people, I had a startup when I was 19, and it was a dry cleaning shop.”

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

What should we make of Jeff Bezos’ decision to move from Seattle to Miami?

It’s tempting to see the Amazon founder’s departure as a blow to the Seattle region. But like many of us, he was a transplant, and different from Microsoft co-founders Bill Gates and the late Paul Allen in that way, in his reasons for being here and ultimately in his long-term commitment to the place. 

That said, Bezos did as much as anyone to change this city and region, indirectly and directly, for better and worse, over the course of three decades. In that way, at least symbolically, it’s the end of an era.

Regardless of his professed reasons for leaving, there are important questions to be asked about what this says about the state’s new capital gains tax structure, and what it will mean to have one of the region’s longtime business leaders and biggest philanthropists living in another state. 

But in the post-pandemic era, when physical location matters less than ever, and when public opinion about billionaires is perhaps lower than ever, Bezos bolting for Miami will cause a lot less hand-wringing in Seattle than Boeing jetting off to Chicago did. He’s still the executive chairman of a major Seattle-based company, after all.

Presumably they’ll let him sleep on a cot at Amazon HQ when he visits for board meetings. 

That’s our first topic on this week’s episode of the GeekWire Podcast. Also on the show this week: Convoy’s final destination; a new AI feature from Seattle startup Yoodli that helps startup founders hone their pitches; and Amazon tests Bluetooth earplugs for warehouse workers.

Related stories and links

• ‘Seattle, you will always have a piece of my heart’: Jeff Bezos leaving Amazon’s hometown for Miami
• Jeff Bezos’ move out of Washington state sparks questions about wealth taxes
• Swimming from Seattle? Jeff Bezos’ big move brings us full circle from the mid-1990s

Flexport buys Convoy’s tech; Convoy CEO Dan Lewis, other employees to join freight company

AI-powered ‘VC pitch coach’ from Yoodli helps startup founders prepare to face investors

Amazon tests letting warehouse workers listen to music and other audio via Bluetooth earplugs

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

This week on the GeekWire Podcast, we’re featuring a panel discussion from the recent GeekWire Summit in Seattle, with three technology and business leaders offering first-hand insights into the new era of artificial intelligence.

• Bridget Frey, CTO at Redfin, the tech-powered real estate company that operates in more than 100 markets in the U.S. and Canada.
• Inbal Shani, chief product officer at GitHub, the software development platform used by more than 100 million developers around the world.
• David Shim, CEO at Read AI, a Seattle-based startup using AI to provide a new window into meetings, and transform them in the process.

Frey, Shani and Shim share real-world examples of AI impacting software development, real estate, and meetings. They address topics like privacy, bias, education, and the future of work. They also discuss the changing nature of technical jobs, and a blurring of the line between developers and non-developers.

From the audience, we get questions about adapting AI to account for emotional intelligence; advice for aspiring engineers; preventing synthetic content from corrupting human experiences; protecting proprietary corporate data; and the prospects for improving work-life balance as AI increases productivity.

Listen above, and keep reading for notes and takeaways.

How Redfin is using AI: “It’s been off the charts,” Frey said. “Really, the amount of innovation and the pace of innovation around AI has picked up so much.” The areas where AI has had the most impact are ones that previously involved humans absorbing and synthesizing large amounts of information.

• Using ChatGPT to generate 4x more localized real estate content.
• Developing new AI-powered features across the company in areas like the home valuation process, including explaining home valuations to customers who ask how Redfin arrived at the number.
• Accepting around 25% of code suggestions from GitHub Copilot, increasing efficiency. Copilot is like a virtual AI teammate that operates alongside developers, automatically generating code and helping them do their work.

The average code acceptance rate for GitHub Copilot is even higher, as much as 55%, depending on the use case, customer and license type, Shani said.

Senior engineers tend to have the most success with Copilot, because they are accustomed to explaining things to more junior engineers, and they can adapt that skill set to Copilot, Frey said. However, younger engineers can play an important role in introducing an organization to the cutting-edge tools they’ve been using in school.

GitHub is now focused on applying AI across the software development life cycle, especially on enabling developers to write more secure code, Shani said. This will come through at the upcoming GitHub Universe conference, she said.

The company also takes responsibility for the next generation of developers in its education program, which has more than 5 million learners. It’s preparing the next generation of developers for a world where AI is table stakes, Shani said.

AI product developers need to be wary of information overload, and up front about privacy issues.

• ReadAI initially provided real-time meeting information, telling speakers, for example, whether people were paying attention to them in the moment. Based on user feedback, it shifted to a post-meeting recap instead.
• The startup notifies meeting participants that recording and analysis is occurring at the outset, and allows users to opt out of the meeting and have all related data deleted by typing “opt out” in the chat, Shim said.

Frey on dealing with issues of bias: “Redfin has 20 years of experience with fair housing law, and the large language models have no experience with it. But we can’t simply reencode years of bias into algorithms and call it a day. So we have been partnering with the large language model companies to share with them test cases and rules that we think need to be handled by these large language models, so that they answer questions that in a way that does not violate fair housing law. We’re very optimistic that we’re going to be able to make real progress on this.”

• Another issue, she said, is making sure that AI doesn’t skew toward pleasing people with its answers at the expense of accuracy.

Social norms around AI are still evolving: “I think the most embarrassing thing is when your AI beats you to the meeting, when you’re five minutes late. That kind of makes you look bad,” Shim said. “People have asked us, ‘Hey, can you make it come in a minute or two after I join?’ “

Shim’s advice for AI startups: “You still have to have your own secret sauce. You need a moat. You just can’t go in and say, I’m another wrapper company … someone who uses ChatGPT, LLaMA, Bard, something else, and just does does prompt engineering. You could actually grow pretty big. There are a number of apps on the marketplace right now that do $100,000 or $200,000 a day in terms of sales, but it drops off pretty quickly because it isn’t differentiated.”

• He notes that around 93% of ReadAI’s processing/engineering resources go into proprietary models, and the remaining 7% is focused on outside models.

Shani says we’re in the Industrial Revolution for AI. “Eventually we’ll find a balance between that niche AI and the generalized AI,” she said. “We’re still working in specific areas of domain expertise. There is only so far that generalized AI can take us, and we will need to build on top of that.”

Each panelist’s concluding message to the audience:

• Frey: “There have been parts of our business that we thought technology couldn’t really change. And just in the last few months, as these technologies have come to light … it’s just created this wave of innovation. So I just encourage you to look through your business, look at parts of your business that you thought technology could never touch. Maybe now there is a way that you could make that part of your business better.”
• Shani: “For us, it’s a lot about developer happiness, which comes from developer experience, developer productivity and developer collaboration. It’s about how we continue innovating and taking AI across the software development lifecycle. And it’s a lot of focus on education across the board … to make sure everyone is ready for the revolution that is happening right now.”
• Shim: “Embrace AI. Go as fast as you can. But make sure you have a moat, because this is going to be a period of time where everything is condensed. Things are moving faster today than they ever have.” Iterate quickly, change as needed with the market, but maintain your core areas of differentiation.”

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.

Podcast audio editing by Curt Milton. On-site A/V by Adavanza.