The biggest figures in the AI field sparred over the risks of the rapidly advancing technology at the World Economic Forum this week, where the buzz revolved around a $500 billion AI infrastructure project promoted by Donald Trump.
AI pioneers, including Google DeepMind chief Sir Demis Hassabis, Anthropic co-founder Dario Amodei, and “Godfather of AI” computer scientist Yoshua Bengio, used the meeting in Davos to repeat stark warnings about AI threats, as business interests and geopolitical rivalries take over. Concerns about safety. .
While Hassabis acknowledged that “it is not possible to put the genie back in the bottle,” he said Artificial general intelligence – When computers exceed human cognitive capabilities – they can threaten civilization if they get out of control or hijacked by bad actors. This is especially the case with large “open source” language models that are accessible to everyone.
“There is much more at stake here than just companies or products,” the Nobel laureate said in an interview with the Financial Times. “(It's) the future of humanity, the human condition, and where we want to go as a society.”
Amodei, whose startup created the chatbot Claude and is Powered by Google And Amazon, he said he was concerned about authoritarian governments using artificial intelligence, and was “very concerned about 1984 scenarios, or worse.”
“Science does not know how we can control machines that are even at our level of intelligence, and even worse, are smarter than us,” Bengio added during one of the sessions. “There are people who say: Don't worry, we will find a solution to the problem. But if we don't find out, do you understand the consequences?
Their stance has been criticized as hypocritical by Yann LeCun, chief AI scientist at Meta, which has spent billions developing open source LLM software called Llama. He said that these fears were belied by the fierce competition between his competitors to build and sell the best models.
“Joshua and Dario have expressed opinions against open source, and this is actually very dangerous,” he said in an interview. “Obstacles to open source distribution may lead to regulatory capture by a few players, either from the US West Coast or China. . . . (put) power in the hands of a few people.
“It's very strange coming from people like Dario. “We met yesterday where he said that the benefits and risks of AI are roughly the same size, and I said: ‘If you really believe that, why do you keep working on AI?’” Lacon added. “So I think he's a little bit two-faced about that.”
While scientists and engineers debated the risks and rewards of AI, business executives showed unrestrained enthusiasm for the technology.
“There are no contrarians,” said Ervin To, head of Dutch technology investment group Prosus. “If you have any appreciation for what big language models and agents trained on them can do, you would be hard-pressed as a human not to conclude that it is transformative and that it will be incredibly disruptive to every industry.”
On Wednesday, the frenetic atmosphere was fueled by OpenAI, SoftBank and Oracle Announcing a joint AI infrastructure project in the United States worth $500 billion It's called the “Star Gate”.
Trump hosted their CEOs, Sam Altman, Masayoshi Son and Larry Ellison, in the Oval Office on Tuesday, before signing executive orders this week that will remove many of the barriers surrounding technology development. The new American president said that these steps will ensure American superiority in technology.
“At OpenAI, we believe infrastructure is destiny,” said Sarah Friar, CFO of OpenAI. “(Stargate) is about more computing. More computing builds better models. Better models answer more complex problems and deliver more benefits to individuals and businesses.
Stargate dominated the discussion in Davos for the rest of the week, with many, including Elon Musk, taking to his social media site It would finance the enormous expenses promised.
The Financial Times reported on Friday that Stargate has not yet received the funding it needs, will not receive any government funding and will only operate OpenAI Once finished. So far, SoftBank and OpenAI intend to provide more than $15 billion each to the project, hoping to raise a mix of equity from their existing backers and debt to finance Stargate.
The new project was also seen as the latest evidence of a rift in the relationship between Altman, Microsoft CEO Satya Nadella and his chief artificial intelligence officer Mustafa Suleiman, the former co-founder of DeepMind who left his startup and joined Microsoft early last year.
“The tensions that emerged between Mustafa Suleiman and Sam Altman in Davos last year were just the beginning,” said Marc Benioff, CEO of Salesforce, which competes with Microsoft in selling AI-powered agents to businesses.
“Microsoft is now accelerating the development of its own AI… This pattern reflects Microsoft’s history with its partners,” Benioff added. “This may mark the beginning of the end for this relationship, making it necessary for OpenAI to expand to other platforms quickly.”
“Mark has no idea what he's talking about,” Microsoft spokesman Frank Shaw said.
Microsoft It has invested nearly $14 billion in OpenAI since 2019 and in return has negotiated rights to its intellectual property and to be the exclusive cloud computing provider. But the last agreement was terminated simultaneously with the announcement of Stargate.
In Davos, Nadella also cast doubt on Stargate spending pledges, touting a planned $80 billion in capital expenditures from Microsoft.
“All I know is that I'm good for my $80 billion,” he later said. Reply Musk said on the social media platform
Stargate is just the latest example of a data center infrastructure arms race in the United States as it prepares for the next phase of the AI economic boom. Musk's xAI is built Supercomputer called “Colossus” It contained 100,000 interconnected Nvidia chips in just three months last year and has pledged to increase the number 10-fold.
BlackRock and Microsoft are preparing to launch a $30 billion artificial intelligence investment fund to build data centers and energy projects to meet the growing demands emerging from the technology sector. Meta President Mark Zuckerberg said on Friday that the company will spend between $60 billion and $65 billion on capital infrastructure this year while expanding its artificial intelligence teams.
“I have ongoing meetings with customers in every sector. I don't think there's a single CEO I've talked to who doesn't know they need to deploy AI,” said OpenAI's Fryar. “AI isn't just on the agenda; it's the agenda. “It is no longer just an abstract concept or a vision of the future. It is here.”
Additional reporting by Harriet Agnew in Davos
