How is artificial intelligence making marketers more productive? How are journalists approaching the ethical dilemmas inherent to AI content creation? And how can scrappy startups hope to compete with well-funded tech giants in the fast-moving world of AI?
Those are some of the questions addressed by tech entrepreneur and marketer Adam Tratt and GeekWire co-founder Todd Bishop on this episode of Shift AI, a show that explores what it takes to adapt to the changing workplace in the digital age of remote work and AI.
Listen below, and continue reading for highlights, edited for context and clarity. Subscribe to the Shift AI Podcast and hear more episodes at ShiftAIPodcast.com.
AI and productivity
Todd: We’ve seen the awakening over the past year. People are trying to figure out what exactly to do with these tools. The capabilities and the weaknesses are becoming more and more apparent. And we’re at the stage where a lot of startups and big tech companies are looking for specific, vertical applications to provide more value.
I was talking with an entrepreneur who said he thinks that each of his engineers has gotten about 20% more productive over the past year to 18 months. And that feels about right to me. That’s about what my experience is
Right now we’re seeing a variety of companies in many different ways look for incremental productivity beyond that, trying to get up to maybe 30%, 40%, 50%. That’s my sense of things right now.
Adam: A lot of the work that I do is by myself. There’s so many different sub-disciplines within the category of marketing, and it’s hard to be great at all of them.
In fact, there’s almost nobody that’s great at all of them, because half of marketing is this analytical, spreadsheet exercise, where you’re trying to make every dollar that you spend work as hard as possible for your brand. It’s usually in driving demand.
Then there’s another part that’s a creative exercise. That has a lot to do with, how do you present your brand? What are the words you choose? What are the visuals you choose?
There’s all these ways to use AI to help you compensate for whatever your weaknesses are. So whether it’s running an efficient LinkedIn advertising campaign, or creating tweets from a blog post, there’s ways you can deploy LLMs and other tools to do a great job.
AI startups vs. tech giants
Boaz: I’m seeing the big companies move so much faster. You wake up in the morning, like we did with Sora and OpenAI, and all of a sudden, all these small companies that were building text-to-video, they are now in a really, really awkward position to have to prove their value.
Todd: The scarcity of computing power is the huge hurdle. So you see this reshuffling of priorities and resources. I think you need some really savvy startups and some really smart founders to navigate this whole world, when you’ve got these giants who are moving so quickly.
Adam: Everybody I know is surprised that Microsoft has come out so strong, so fast. And Google is still apparently flat-footed. It’s very early days. And I would not be surprised if both Google and Amazon come out with very competitive offerings, as the smoke starts to clear. We all know Google has been working on AI for a long time. And they were intentionally pumping the brakes, until they realized OpenAI was eating their lunch.
[Editor’s Note: This podcast was recorded before Apple’s recent AI announcements.]
Todd: My impression walking away from Google I/O was that Google CEO Sundar Pichai is far more conservative, in terms of pushing the envelope on AI, than certainly either OpenAI CEO] Sam Altman, or Microsoft CEO Satya Nadella. My question walking away from that was, what has he seen?Adam: It might just be that he’s seen the Terminator trilogy.
The ethics of AI content creation
Adam: I do have some ethical concerns, but the way I’m using AI is, at the moment, somewhat benign: “Read this blog post, tell me what it says, help me create provocative, interesting social media posts, take this story, create an image for the story.” … I have no ethical qualms with generating a fake image for a story, as long as it doesn’t misrepresent the thing that I’m trying to say, and you need to credit appropriately.
Todd: That brings up an interesting ethical dilemma that we ran into at GeekWire. Recently, we decided, when we’re generating images to go with stories, we will not use photorealistic images. We’ll use things that are clearly illustrations, created by the AI generator, because otherwise there could be this confusion in the minds of readers. So we’re just coming up with these things on the fly — things that feel right based on our underlying principles.
Two words about the future of work and AI
Adam: Guarded optimism. In general, what I’ve seen in my career in tech has been steady progress, mostly for good. I think the people I’ve worked with have had good intentions. Occasionally, in the course of being competitive, things get out of hand. And then usually the government steps in, too late. I’m hoping that, just as we’ve seen in the first year-and-a-half of general availability of LLMs, that the promise will be elusive compared to the reality, and we’ll have time as an industry to course-correct as we get closer to this idea of more sophisticated models.
Todd: Hyper drive. I think that things are going to just accelerate and continue accelerating. I’ve been covering tech since 2002, more than two decades. I’ve never seen a year like this, ever. I missed the dot-com bubble. I didn’t get to see that. And I feel like this has made up for it in some respects. I just feel like it’s going to accelerate so fast that the only thing that’s going to slow us down is a backlash.
Listen to the full episode of Shift AI with Todd Bishop and Adam Tratt here.
