The AI startup world may be getting hotter, but there’s a difference A startup using artificial intelligence and AI-First Startups – The Allen Institute for Artificial Intelligence cultivates the latter in its AI2 IncubatorTheir previous $10 million fund has done so well that backers have returned and refilled the coffers three times over.
Since 2017, 21 companies have come through the incubator, attracting about $160M in further investment and at least one major acquisition: XNOR, an AI acceleration and efficiency agency, followed (I think this is a direct result of the TC report) Result) was snapped up and acquired by Apple for about $200 million.
XNOR’s success itself may be enough to justify a second fund, but AI2 has the ability to find interesting companies with tech founders chasing undervalued markets. Two come to mind: WellSaid has been providing synthetic speech services for a long time, and Blue Canoe uses ML to help people deal with foreign language accents.
“We’re at an extremely exciting time in the AI world, but the lesson is to stay calm and be pragmatic in the face of extraordinary hype. It’s easy to drink too much Kool-Aid,” said Jacob Colker, the group’s managing director. “We continue to beat the drum of entrepreneurial fundamentals: solving real problems, creating real experiences, providing firepower to teams and developing go-to-market strategies, and we’re doing it with very talented techies.”
The technical prowess of their founders is a deliberate move by the incubator and AI2 in general.
“We’re specifically looking for technical founders because they might need to build their own version of the model, or run a million custom models instead of one big model. It comes down to our roots: the AI2 team, basically 50% of our management People are highly, highly technical,” said Oren Etzioni, technical director at AI2.
The team is careful to emphasize that the goal is not to produce some dark horse capable of challenging big companies and models, or to simply capitalize on the current frenzy around AI following the historic popularity of ChatGPT.
“A lot of companies are marketing on top of OpenAI’s models, but you need to do more than just put lipstick on the API,” Etzioni said. “And we’re not trying to build a language model that’s better than OpenAI — let the multibillion-dollar company decide. But what if I can make a model that’s good for accounting or public speaking? We see huge Space, global problems that are not being adequately addressed, OpenAI is not for it, and someone can build a great company in it.”
“We do encourage our entrepreneurs to step up the pace, take advantage of tailwinds, move faster and raise capital faster because the competition isn’t sitting still,” said Vu Ha, AI2’s technical director, but they are settling for the long haul. “In terms of future potential, the recent wave of advances in AI is even more exciting: New markets and opportunities will be created over the next two years, and no one organization will stand alone.”
Colker highlighted Comwave as an example of AI entering a field that big companies are unlikely to even be aware of, let alone pursue as a business opportunity: nurse burnout. Over the past few years in particular, nurse turnover has been much higher than normal, partly due to the inability at the hospital level to intelligently manage the workload of its staff. Comwave aims to provide an early warning of this, but of course “no one buys a bunch of AI on faith,” Etzioni admits.
As such, the incubator has been guiding companies through the challenges of research, certification, and pilot studies that any healthcare software vendor needs to grapple with. AI2 has a strong network – one of the most important aspects of any such program – which provides an excellent knowledge base for difficult issues, such as entering regulated or technically conservative markets.
Currently, the AI2 incubator is targeting (and exceeding) a pass rate of 5 companies per year, although as Ha says, “With the new funding, we do have the capacity to deploy at a larger scale.” Still, the team believes there is no Necessary to expand beyond seed companies: their expertise is at an early stage, and of course A rounds do tend to be more expensive. $30 million is a lot of money, but when AI companies are raising hundreds of millions every few months… it’s best to stick with what you know.
The fund is backed by Madrona, Sequoia Capital, Vinod Khosla, Evergreen Ventures and “other leading limited partners.” You know who you are!