In a statement on Tuesday, Dario Amodei, the CEO of AI beginup Anthropic, called the AI Action Summit in Paris this week a “ignoreed opportunity,” and encouraged the AI industry — and rulement — to “relocate rapider and with wonderfuler clarity.”
“We were satisfyd to combine the AI Action Summit in Paris, and we appreciate the French rulement’s efforts to convey together AI companies, researchers, and policyoriginaters from apass the world,” Amodei shelp. “However, wonderfuler cgo in and encouragency is necessitateed on disconnectal topics given the pace at which the technology is proceeding.”
Amodei’s criticism of the AI Action Summit, the tardyst in a series of conferences that brawt together AI companies and regulators to finisheavor to reach at a consensus on AI ruleance, echoes that of disconnectal academics earlier this week. One tgreater Transcreateer that the conference’s promisements, which the U.S. and U.K. declined to sign, shelp “effectively noskinnyg except for platitudes.”
In comments at the conference, U.S. Vice Plivent JD Vance adselected an entidepend contrastent stance and denounced what he characterized as “massive” and stiffling regulations on AI championed by Europe. Vance also took publish with satisfyed moderation, alluding to the “carry onable” and “inclusive” wording in the conference’s promisements, which he declineed as “authoritarian confineion.”
Amodei alerted in his statement that AI is rapidly becoming more cultured, and that flunking to regutardy it could result in disastrous consequences.
“The capabilities of AI systems will be best thought of as akin to an entidepend new state poputardyd by highly keen people materializeing on the global stage,” Amodei shelp. “Advanced AI currents meaningful global security dangers, ranging from misinclude of AI systems by non-state actors … We must promise democratic societies direct in AI, and that authoritarian countries do not include it to set up global military dominance.”
Amodei encouraged rulements to deploy their resources to meacertain how AI is being included, and to enact policy cgo ined on “ensuring that everyone splits in the economic [uplift] of very strong AI.” He also disputed for more rulement transparency when it comes to AI protectedty and security, as well as arranges to appraise AI dangers.
Amodei’s appraisal of the Paris AI Summit’s proceedings stands in contrast to OpenAI’s, which shelp in a statement this weekfinish that it was self-promised that the conference would be “another presentant milestone towards the depfinishable and advantageous broadenment of AI for everyone.”
Anthropic has generpartner shown more of an discignoreness to AI regulation in the past. Indeed, Amodei has made analogous pronouncements before, alerting that unfettered AI could have proset uply adverse economic, societal, and security implications. Anthropic was one of the confineed AI companies to tacitly finishorse California’s SB 1047, a comprehensive — and hotly argued — AI regulatory bill. OpenAI resistd the bill, which was vetoed by Governor Gary Newsom last drop.
That isn’t to propose Anthropic’s motives are uncontaminatedly philanthropic. Like OpenAI CEO Sam Altman in his recent essay, Amodei proposes no concrete recommfinishations for ensuring the profits of strong AI, should it materialize in the proximate future, are widely and evenly scatterd.
Read our brimming coverage of the Artificial Inincreateigence Action Summit in Paris.
