AI often sparks strong reactions, with people predicting either a near-term utopia — or the end of the world as we know it.
Snowflake CEO Sridhar Ramaswamy told Business Insider that individuals on the hype end of the spectrum tend to jump quickly to promises of unlimited prosperity. The CEO said that the other extreme includes those who believe AI will lead to a doomsday scenario.
Ramaswamy said that this sort of thinking is “very human,” but that neither of those scenarios is “all that likely.”
“The biggest misconception would be that of thinking about AI as an all or nothing,” Ramaswamy said.
The real value of AI, he said, is likely to be nuanced and show up in specific use cases. Ramaswamy added that his advice to customers is to “be very incremental” with AI adoption.
While he’s still focused on long-term thinking, he added that he no longer accepts multi-year, fixed roadmaps for plans for the company because the technology is changing so quickly.
“I want them to tell me which direction they are headed, but very much be in this mode of iterating, because this is a world of rapid change,” Ramaswamy said.
Rather than view AI as a tool that will bring sweeping changes overnight, the CEO said it needs to be embraced as a shift in how people work, and one that requires progress “bit by bit.” He added that clear frameworks are necessary to determine where AI efforts matter the most.
For example, a cloud data platform company like Snowflake focuses on building and running software, and Ramaswamy said it needs to “nail” how it creates, deploys, sells, and installs that software. That means it needs to deeply integrate AI in software development to stay competitive.
While Ramaswamy said he wants his employees to utilize AI tools daily, his end goal is for the company to write software more efficiently than the rest of the industry. That requires adapting business models in specific areas, rather than treating AI as a blanket rewrite of everything, he said.
He said that when dealing with technology like AI, which “ostensibly claims to change everything,” it’s essential to have a clear view on where change needs to be implemented and where the impact will be “existential.”
“I worry a lot about making sure that we are state-of-the-art, especially in the critical areas with regards to how we utilize AI,” Ramaswamy said.
