When Nick Frost was attending college more than a decade ago, he worried that he was a little behind in the field of AI.
Frost, co-founder of enterprise AI startup Cohere, said during a recent episode of TechCrunch's Found podcast that he came to that conclusion in 2012 after Geoffrey Hinton published research showing that neural networks could be trained to successfully identify objects like cars and animals. Looking back, Frost said that while that research is archaic compared to current AI capabilities, at the time he felt like he'd missed out on a technological breakthrough.
“It felt like a missed opportunity,” Frost said. “I remember thinking at the time, 'Oh my God, if only I'd started college a few years earlier, I'd have been there from the start.'”
Of course, he didn't delay and founded Cohere in 2020 to build custom AI models for enterprise customers. The company has raised over $934 million in venture capital and is currently valued at $5.5 billion.
Frost spoke about why he and his co-founders left Google to start Cohere, and how research done by current co-founder Aidan Gomez on why general-purpose AI models are better than more specialized or verticalized models was fundamental to the approach to building Cohere.
“We're not trying to build a consumer product like our competitors, and we're not developing 1,000 different things at once,” Frost said. “We're trying to build language models that are really useful for enterprises, but that single focus isn't something that can be built inside a huge multinational corporation.”
Frost also spoke about why he thinks the AI industry shouldn't shy away from tough questions around things like regulation and sustainability, and why he's glad the industry is becoming more realistic about what AI technology can and can't do.
“I don't think we'll get to artificial general intelligence, I don't think we'll get to superintelligence, I don't think we'll get to digital gods anywhere in the near future,” Frost said. “I think more and more people are waking up to it and saying this technology is great, it's very powerful and it's very useful. [but] “It's not a digital god. We need to adjust how we think about technology.”