This is a fragment Sources: Alex Heatha newsletter about artificial intelligence and the tech industry, distributed once a week only to The Verge subscribers.
Leaders of three leading pioneering artificial intelligence labs spent this week at the World Economic Forum in Davos, Switzerland, taking shots at each other like candidates in a presidential primary.
I helped start the news cycle. During Tuesday’s conversation, among others: I asked Google DeepMind CEO Demis Hassabis on OpenAI’s decision to test ads in ChatGPT. “It’s interesting that they decided to do it so early,” he said. “Maybe they feel they need to earn more.”
The next day, Anthropic CEO Dario Amodei shared this information during an interview I watched at The Wall Street Journal House in Davos. “We don’t need to make money off a billion free users because we’re in a death race with some other big player,” he said. He also teased an upcoming essay focusing on the “bad things” that AI might bring – a murky counterpart to his optimism “Machines of Loving Grace” Essay. from last year. During another appearance in Davos, he compared the U.S. allowing Nvidia to sell graphics processors to China to “selling nuclear weapons to North Korea.”
“We don’t need to make money off a billion free users because we’re in a death race with some other big player.”
OpenAI’s response came from Chris Lehane, the company’s head of policy and perhaps the most hazardous political operator in Silicon Valley. Lehane earned the nickname “disaster master” in the Clinton White House, where he specialized in opposition research and crisis management. At Airbnb, he helped the company weather regulatory battles that threatened its existence. He is now the most senior policy chief at any AI lab, and has employed tactics from the campaign trail to the AI race.
When I sat down with Lehane for breakfast near Davos’ main promenade on Thursday morning, he was ready to answer. In response to comments on Hassabis’ ads, Lehane pointed out the obvious irony. “If you want to give people access, you have to pay for the computation,” he told me. “I’m thrilled to be able to have this conversation with the largest online advertising platform in the world every day, seven days a week.” He also called Amodea’s comments “elitist” and “undemocratic.”
“It’s common for someone trying to move up from the second tier to say provocative things because it creates a feedback loop,” he told me between bites of scrambled eggs. “It attracts attention. In my experience in politics, it’s often short-lived because ultimately, if you say things like that, people will hold you accountable for actual solutions. If we lose a lot of jobs [to AI]what do you actually do to address it, especially if you ask these questions, right?”
“People making this criticism often don’t focus on how to make this technology widely available,” he continued. “They tend to come from backgrounds that focus almost exclusively on enterprise employ cases. It’s a very elitist approach.”
“It’s common for someone who’s trying to move up from the second tier to say provocative things because it creates a feedback loop. It attracts attention.”
The reality is much more nuanced than the attacks that AI labs inflict on each other. OpenAI is aggressively trying to take over Anthropic’s enterprise AI business, as is Google. And while it’s true that ChatGPT is the most popular chatbot, turning its advertising push into an element of some democratic virtue rather than a financially motivated move to ultimately monetize most of ChatGPT’s usage is quite a diversion.
Throughout our conversation, Lehane kept returning to his political framework. He told me that being in Davos “felt a bit like walking through downtown” Manchester, New Hampshire, before a major race: the weather, signs everywhere, and campaigns converging in one compressed environment, all trying to get attention.
“We have leadership status,” Lehane said. “Even if the leader started as a murky horse, we have now established ourselves based on our innovation. And the rest are trying to counter it.”
After talking to AI leaders this week in Davos, I got the impression that the industry has collectively decided to join forces on OpenAI. Hassabis and Amodei praised each other on stage during this week’s official WEF panel titled “The day after AGI.”
“I think what we really have in common is that both companies are run by model-focused researchers who are focused on solving important problems in the world,” Amodei said during the panel. “I think these kinds of companies will be successful in the future, and I think we have a lot in common.” (Altman himself left Davos this year, as it stands apparently in the Middle East, raising tens of billions of dollars more.)
Meanwhile, OpenAI’s rivals have told me they are particularly annoyed by Altman’s aggressive attempts to shore up AI’s potential, and some are frustrated at being kicked out of the deal by an unprofitable company that hasn’t yet demonstrated it has enough revenue to cover the impressive commitments it’s making.
With hundreds of billions of dollars and the race to AGI accelerating, I expect the rhetoric to get more heated this year. Lehane told me that as Election Day gets closer, the campaigns get nastier. If he’s right about the analogy, we’re still early in the primaries.