Anthropic agrees with the U.S. authorities that implementing sturdy export controls on domestically made AI chips will assist the U.S. compete within the AI race towards China. However the firm is suggesting just a few tweaks to the proposed restrictions.
Anthropic launched a weblog submit on Wednesday stating that the corporate “strongly helps” the U.S. Division of Commerce’s “Framework for Synthetic Intelligence Diffusion” forward of the interim rule’s implementation date on Might 15.
The framework was proposed by outgoing president Joe Biden in January and is supposed to bolster AI chip export controls for the needs of nationwide safety and to make sure the U.S.’ dominance in AI. It divided the world’s international locations into three tiers, with every tier having its personal tips and restrictions.
Tier 3, probably the most restrictive tier, which incorporates international locations that have been already impacted by present export controls, like Russia and China, would face further restrictions. Tier 2 international locations, like Mexico and Portugal, would come beneath export restrictions for the primary time and would have a cap on what number of chips they might purchase. Tier 1 international locations, like Japan and South Korea, would proceed with out export restrictions.
When these restrictions have been proposed in January, semiconductor large Nvidia launched an announcement calling them “unprecedented and misguided,” and suggesting that they’d “derail” innovation worldwide.
Clearly, U.S.-based AI corporations, like Anthropic, don’t agree. In its weblog submit, the lab expressed assist in broad strokes for the restrictions.
Anthropic did, nonetheless, suggest decreasing the variety of chips Tier 2 international locations should buy with out evaluate and as an alternative encouraging these international locations to purchase extra chips via government-to-government agreements to keep away from smuggling and improve U.S. management.
Techcrunch occasion
Berkeley, CA
|
June 5
BOOK NOW
The corporate additionally thinks the U.S. authorities ought to improve funding to make sure these export controls are correctly enforced.
This assertion by Anthropic isn’t notably stunning. Anthropic’s CEO Dario Amodei has been one of many extra vocal U.S. AI leaders in favor of export restrictions. Amodei wrote an op-ed within the Wall Avenue Journal in January about why the U.S. wants stronger chip export controls.
iinfoai has reached out to Anthropic for extra data.
