As a devastating fire raced through Lahaina last summer, planning for emergency aid and recovery was well underway, even before aid organizations could arrive. Aid workers knew just where to go in the ravaged town to make best use of their resources, thanks to quick work at Microsoft’s philanthropic AI for Good Research Lab. Launched in 2018, the Lab shares its AI models with nonprofits that lack the resources to build their own.
Employing damage-assessment techniques, the Lab’s nearly 40-person team had already road-tested in the Ukraine war and last February’s earthquake in Turkey. In each case, the Lab’s AI model needed just a few hours to review satellite data and map the damage levels of every affected building. Prior to employing AI for this sort of disaster-response, aid personnel would spend weeks assessing the situation on the ground, slowing the response effort and potentially putting workers in danger.
“It’s a game-changer to have a very good understanding of the situation before you arrive,” says Juan Lavista Ferres, chief scientist and Lab director.
As recently as last fall, Seattle didn’t seem to register as a major AI hub on many observers’ radar, despite being home to top-notch institutions including the Allen Institute for AI. But that’s changing fast, as major companies in the region explore high-profile projects that use AI to tackle global problems.
Two of the top-ten holders of AI patents, Microsoft (over 2,300) and Amazon (over 1,000), are based in the Seattle area. Other major AI players Meta, Google, and Apple all maintain major engineering centers here as well. The region was recently ranked 4th in the nation for AI and machine learning innovation by commercial real-estate tracking firm Commercial Café.
Investors are also taking more notice of our region’s AI startup community. In the fourth quarter of 2023, ten AI startups based in the Northwest raised capital that totaled nearly $105 million.
The biggest round in the Seattle area was closed by startup Rhythms, which came out of stealth mode in December ’23 to launch its team-productivity AI system with a $26 million “seed” round led by Madrona Ventures. AI-watchers are keeping an eye on this one, as founder Vetri Vellore sold his previous AI startup, Ally, to Microsoft in 2021.
Besides its disaster-response mapping tool, Microsoft’s AI for Good team is pursuing land mapping for fire-danger forecasting and disease-outbreak prediction. Health-focused projects include enhanced breast-cancer and prostate-cancer detection. With AI models that have been trained for months to analyze patterns in MRIs for possible discovery of new cancer markers, doctors ¬– with the AI tools’ assistance – can potentially catch more cancers in their brief one- or two-minute review of a scan, Lavista Ferres explained.
Where Microsoft’s AI initiatives are manifold, local AI startups typically seek to solve a single problem. ioCurrents co-founder and chief innovation officer Cosmo King grew up on the water in Virginia, spurring his interest in how AI could help the maritime industry. The Kirkland-based startup has developed an AI model that helps commercial ships cut fuel costs, optimize maintenance schedules, and reduce equipment breakdowns.
In researching the issue, King found ships today use far more technology than they used to, making it impossible for crew to constantly monitor every gauge and indicator.
“They’re basically floating data centers with 200 sensors on every engine, where in the past there might have been five,” he said. “But there was no good real-time machine data available.”
The company’s AI model assesses all of a ship’s data, comparing it with data from similar vessels to find ways to reduce fuel usage. There’s growing interest in the industry in fuel efficiency, both for the cost-savings and to meet requirements to reduce carbon emissions. This makes ioCurrents’ solution more of a ‘gotta have’ than a ‘wanna have.’
“We had to have compelling return on investment cases for this industry, or it was a non-starter,” King said. “They don’t just buy cool tech–they’re very busy and cost-focused.”
Current customers include Hornblower, the founding operator of NYC Ferry, which is committed to improving fuel efficiency, said King. Tugboats nationwide also use ioCurrents’ solution, a market segment King said he hadn’t anticipated. Other customers are in the Shipping, Offshore Wind, and Offshore Oil and Gas segments around the globe including Africa, Europe, the Americas, and Asia.
As home to the nation’s largest fishing fleet and ferry system, Puget Sound is the ideal home base for maritime-focused ioCurrents, King said. He also helps connect the region’s tech talent with the maritime industry as a board member of Maritime Blue, a regional nonprofit strategic alliance focused on fostering startups in the local maritime sector.
“The maritime industry overlaps with the tech industry here in a completely unique way,” he said. “It’s one of the best places to find talent and maritime is fundamental here.”
While the range of projects may feel like a leap — from cutting carbon emissions on ships to better detection of cancer, AI models can help with both and much more. In fact, Lavista Ferres and Microsoft Director of AI for Health, William Weeks launch a book later this spring entitled, “AI for Good: Applications in Sustainability, Humanitarian Action, and Health.” One of Lavista Ferres’ favorite AI for Good projects will be featured there; a collaboration with Fred Hutchinson Cancer Center recently featured in an AI-focused series with Trevor Noah on QuitBot, a chatbot now in testing that provides support to people trying to kick the nicotine habit.
“If you could only do one thing in life and you helped people stop smoking, that would significantly reduce cancer around the world,” he said. “These are things of importance.”