AI leaders are sounding the alarm, calling for regulations to prevent the unchecked growth of artificial intelligence. The push for regulation comes as the technology continues to advance at a breakneck pace, with some experts warning that AI could become a threat to humanity if left uncontrolled. Dario Amodei, a leading AI researcher, has proposed a set of standards that would ideally be enforced through regulation, targeting all US frontier AI companies that don't voluntarily comply.
However, achieving such regulation may be difficult under the current US administration. President Trump has expressed skepticism, suggesting that the only control AI needs is a 'strong and smart' president. Speaker of the House Mike Johnson has acknowledged the need for safety measures, but has cautioned against panic and urged against imposing emergency moratoriums. David Sacks, co-chair of the president's Council of Advisors on Science & Technology, has urged frontier model makers to self-regulate, rather than waiting for Washington to act.
Amodei's proposal includes a number of different levels of international agreements on AI safety, with the highest being a 'full pacing, or even 'pause,' in which participating governments agree to substantially limit the overall rate of AI development.' However, Amodei acknowledges that such an agreement is unlikely to happen anytime soon, especially given the high stakes and the potential for China to defect and gain a geopolitical advantage.
The issue of China's role in AI development is a major concern, with the country's open-weight models only a few months behind those of major corporate AI behemoths. Amodei's proposal highlights the need for international cooperation to prevent a collective action problem, in which individual countries or companies prioritize their own interests over the need for safety and regulation.
Cet article a été rédigé avec l'assistance de l'IA.
News Factory APP - actualités agentiques pour booster votre SEO et AEO.