Prominent venture capitalist Bill Gurley was pacing the stage at an event this week, asking the audience to shout out a sentence that wouldn’t normally elicit excitement. Gurley, however, received an enthusiastic response from the audience.
“Regulation is the incumbent’s friend!” they shouted.
Gurley was speaking at the All-In Summit in Los Angeles, a largely tech-focused event All podcasts.His speech was titled “2,851 Miles,” which is the distance between Silicon Valley and Washington, D.C.
Gurley, a general partner at venture capital firm Benchmark who has invested in companies such as Uber, Grubhub and Zillow, warned of the dangers of “regulatory capture.” He describes his experience opposing it while supporting innovative startups, and then warns of its role in today’s artificial intelligence landscape.
To explain the concept, he quoted 1982 Nobel Prize winner in economics George Stigler: “Generally speaking, regulation is acquired by industry and is designed and operated primarily for the benefit of industry. ” In other words, special interests take precedence over the general interest of the public.
Gurley recounted his experience at Tropos Networks, a company invested by Benchmark. He described how mayors were initially excited about the company’s wireless mesh technology and wanted to use it to provide municipal Wi-Fi service.
“There are hundreds of mayors across the country who want to provide free Wi-Fi service in their downtown areas,” Gurley said. “This will help with public safety, economic development and, of course, the digital divide.”
Unfortunately, he said, the idea “conflicts with business interests,” namely incumbents and powerful lobbyists. In Philadelphia, Verizon and Comcast used lobbyists to push bills through the Pennsylvania Legislature to protect their positions from upstart challengers like Trobos, he said. Soon, more such regulations spread to other states.
Regulatory Capture Risks of Artificial Intelligence
Gurley presented some other examples of regulatory capture before highlighting a more relevant case today: Artificial Intelligence
he shared on screen New York Times The May article, titled “OpenAI’s Sam Altman urges AI regulation at Senate hearing.”
“Sam is just getting started,” Gurley said, referring to OpenAI CEO Sam Altman. “He also wants regulation.” OpenAI, the maker of the artificial intelligence chatbots ChatGPT and GPT-4, is widely seen as well ahead of its competitors.
“There’s a very scary thing about artificial intelligence,” Gurley said. “The incumbents who are meeting with the administration are spreading something that I don’t think is accurate or fair: they’re spreading negative open source information, and I think that’s precisely because they know that’s their biggest threat. ”
If the large language models (LLMs) that power AI chatbots like ChatGPT were open source, more startups would be able to innovate and challenge incumbents. In contrast, LL.M.s at OpenAI and Google (and its ChatGPT competitor Bard) are generally shielded from public scrutiny.
Tesla CEO Elon Musk co-founded OpenAI but has since stepped away from it, Tweet February: “OpenAI was created as an open source (that’s why I named it “Open” AI), non-profit company to serve as a counterweight to Google, but now it has become a closed source, profit maximizing of companies controlled by Microsoft. Not what I want at all.”
Altman and Microsoft have both denied the claims, while OpenAI’s chief scientist and co-founder Ilya Sutskever shared his thoughts on the reasons for abandoning open source in an interview with The Verge in March:
“We were wrong. Frankly, we were wrong. If you believe, as we do, that at some point artificial intelligence (AGI) will be incredibly powerful, there’s no point in open source. It’s a bad The idea… I fully expect that in a few years it will be completely clear to everyone that open source AI is not a smart idea.”
“At some point, these models are going to be very easy to do a lot of harm if someone wants to,” he added. However, he also noted that “the safety aspect is not as important as the competition aspect yet” and that “there are many, many companies that Want to do the same thing.”
Altman himself told lawmakers in May, “We don’t want to slow down small startups. We don’t want to slow down open source efforts,” while adding, “We still need them to follow the rules.”
Marc Andreessen, a general partner at venture capital firm Andreessen Horowitz, opposes regulatory capture in the AI field, warning in June that “if regulatory barriers are erected, a government-backed cartel of AI vendors will form,” Chief Executive Marc Andreessen said in a statement. Executives will make more money” from new startups and open source competition. ”
Gurley said Llama 2 from Meta, one of the leading open source LL.M.s., “is actually pretty interesting.”
Silicon Valley celebrities including Andreessen, YCombinator co-founder Paul Graham and Greylock partner Reid Hoffman signed a statement of support for Llama 2, which reads as follows:
“We support an open innovation approach to AI. Responsible and open innovation involves us all in the AI development process, bringing visibility, scrutiny and trust to these technologies. Opening the Llama model today will allow everyone to benefit from this benefit from this technology.”
Near the end of his speech, Gurley warned, “If you care about prosperity and stifle innovation, you’ll stifle prosperity.”
He concluded his speech by referring to the title “2,851 Miles.”
“The reason Silicon Valley is so successful,” he said, “is because it’s so far away from Washington, D.C.