OpenAI’s Sam Altman on what’s happening to Silicon Valley

OpenAI CEO Sam Altman says Silicon Valley needs to give research groups more freedom. David Paul Morris/Bloomberg via Getty Images

OpenAI CEO Sam Altman is undeniably a Silicon Valley insider. In addition to leading the company behind the AI ​​chatbots ChatGPT and GPT-4, he also leads the startup accelerator Y Combinator.

But even though he was part of Silicon Valley, he wasn’t satisfied with it.

“I say this because I realize it will come across as arrogant and I don’t mean it that way,” he said on Wednesday’s episode of the show In good company audio file. “There has been great research done in Silicon Valley companies…There hasn’t been anything like that in a long time.”

Podcast host Nicolai Tangen, CEO of Norwegian sovereign wealth fund Norges Bank Investment, expressed surprise at this stance.

Altman replied that there is good product innovation in Silicon Valley, but asked rhetorically: “Before OpenAI, what was the last truly great scientific breakthrough to come from a Silicon Valley company? ?”

As for why the culture behind those breakthroughs disappeared, he said: “I spent a lot of time pondering that question. I don’t completely understand it.”

One reason, he says, is that today’s tech companies don’t give their research teams enough freedom. At OpenAI, by contrast, “we set a very high vision for the company and what we want to achieve, and beyond that, the researchers get a lot of freedom,” he said.

He noted that researchers are allowed to explore different directions. When the most promising became clear, OpenAI was able to get “nearly the entire research brain trust behind it,” eventually leading to the release of ChatGPT last year, which helped spark the boom. about artificial intelligence today.

One problem in Silicon Valley, he said, is that “it has become so easy to create a super valuable company and people have become so impatient about timelines and payback deadlines that a lot of capital is Pouring into these things can, you know, multiply money quite reliably in a short period of time…That has attracted a lot of talent, understandably.”

Most big tech companies, he continued, started as a product company and ended up with “a research lab that didn’t work very well.” In contrast, OpenAI started as a research lab.

In March, Altman said he was “seriously questions” the advice he has given startups over the years at YCombinator, noting that there are various ways in which OpenAI “goes against all of YC’s advice.”

“It took us four and a half years to launch a product,” he said at the Stripe conference. We will become the most capital-intensive startup in the history of Silicon Valley. We were building a technology without knowing who our customers would be or what they would use it for.”

Greg Brockman, president and co-founder of OpenAI, has also noted the company’s contrasting approach. “You should have a problem to solve, not a technology looking for a solution,” he said above Feasibility podcast in March.

Of course, OpenAI has benefited from Microsoft investing billions of dollars in it, a luxury that very few startups have.

As for the future, Altman told Tangen that he and OpenAI have gone beyond thinking about a roadmap to follow. “We’re just doing a bunch of things that are outside the norm that Silicon Valley has received. And so we just say, ‘Oh, we’ll figure it out, and we’ll try things, and if we get it wrong, like, who cares…it’s not like us ruined something that was already figured out.’”