ADVERTISEMENT

Technology

AI companies focusing on production over promises in 2024: Cohere co-founder

How AI could change the global tech sector Nick Frosst, co-founder at Cohere, joins BNN Bloomberg to discuss the real life impact of AI and the changes it could bring to tech.

At the Collision conference in Toronto this week, artificial intelligence pioneer Geoffrey Hinton warned the North American technological community that surveillance risks are one of the most urgent problems posed by AI.

This, alongside Hinton’s doomsday-like concerns for artificial intelligence including lethal autonomous weapons, corrupted elections and job losses, are not outcomes Canadian enterprise AI company Cohere feels are synonymous with growing out the technology.

In an interview with BNN Bloomberg on Wednesday, Nick Frosst, co-founder of Cohere says the science-fiction-styled doomsday scenario some fear will emerge from AI is not realistic.

“Geoffrey is a dear friend of mine. I started my work here as a researcher in his lab and I know pretty much everything I know about machine learning research from him, so we continue to debate this frequently but we have very different points of view,” Frosst says.

“At Cohere we are an enterprise-focused, large language model provider. We don’t think this technology presents a doomsday scenario. We do think there are real-world risks to using this technology and we take them very seriously,” he says, noting data privacy and security as top priorities.

“We (Hinton and Frosst) have been debating it for many years now and I think we have influenced each other a little bit, but I think over the years we’ve stayed pretty divergent on this viewpoint,” notes Frosst.

Real-world impact of AI

While AI developers debate the philosophical implications of expanding the technology, global investors are waiting to see if the hype is translating into tangible assets; Cohere’s Frosst is optimistic.

Last year was the year of showing the world what AI could do, this year is a year of production, he says.

“We are starting to see a lot more companies go into production,” Frosst says, “but getting to production requires solving a lot of problems that have nothing to do with AI, some of them are security issues or user experience issues and those all take time.”

For Cohere, artificial intelligence will impact the business community similar to the invention of the touch-screen, the personal computer or the internet, rather than more niche inventions like 3D printing.

“There was a lot of hype. There are some things large language models can’t do. They can’t predict the future but there are a lot of business applications they can do that are indeed in production already, solving problems,” said Frosst.

“It will be commonplace, a part of our lives and influence the way we do work but it will also be just part of our everyday existence at some point. In some ways, the world will look similar but significantly more effective and efficient.”

Cohere’s competitive strategy

At a time when big tech companies are competing for the top spots in AI leadership, Cohere is focusing squarely on enterprise applications of the technology, something he sees as a key difference between his company and San Francisco-based AI-giant OpenAi.

“We are exclusively focused on solving business problems and building solutions for enterprise, so we are not interested in consumer applications,” Frosst said adding that led the company to build various “real world things” for a range of customers.

In the past four months, Cohere has launched four new initiatives: two AI models named Command R and Command R+, as well as launched a developer toolkit and a startup program for early-stage companies.

“It has been a really exciting start to the year for us,” says Frosst.

Canada’s capital gains tax impact

Canada’s new capital gains tax is set to come into effect as of June 25  this year, but Frosst says this change is not swaying Cohere's commitment to Canada.

“We are proud to be Canadian, we chose to grow the company here. We have no plans of changing that in the future,” he says. “For machine learning and neural networks, there’s a strategic advantage to being here (in Toronto). It’s a really great place to build an AI company.”

“We would love to see the government continue to support innovation and continue to support businesses,” he adds.

Frosst says his company has no plans to go public. “For now, we’re happy to be a private company. We are in a good position.”