Madaxweynaha Anthropic Dario Amodei wuxuu soo saaray qorshe lagu xakameynayo horumarka AI
Dario Amodei, hoggaamiyaha shirkadda AI-ka ah ee Anthropic, wuxuu maanta soo saaray blog uu ku sharxay saddex tallaabo oo lagu xakameynayo horumarka AI, isagoo shirkaddiisa ku dhawaaqay inay si toos ah u qabsoonto mid ka mid ah tallaabooyinkaas.
Tallaabooyinka ugu muhiimsan ee lagu xakameynayo AI
Amodei wuxuu sheegay in shirkaddiisu ay si toos ah u qabsoonto inay u diyaariyo 'baraysiyeeyaasha dhexdhexaadka ah' (embedded evaluators) oo ka socda hay'ado saddexaad sida METR, kuwaas oo la wareegaya baaritaanka amniga AI-ka. Wuxuu sidoo kale codsaday in shirkadaha ugu horreeya ee AI-ka ee dalalka dimuqraadiyadda ku jira ay isku raacaan 'xisaabinta amniga iyo xad-dhaafka ah' ee horumarka AI-ka.
Wuxuu intaas ku daray in dowladda Mareykanka ay u baahan tahay inay ka qayb-gasho wadahadallada amniga, laakiin aanay u baahnayn inay ka qayb-gasho, iyadoo kaliya u baahan tahay inay bixiso 'ka joojin' (waiver) gaaban oo ku saabsan wadahadallada amniga. Wuxuu sheegay in tallaabooyinka sida diidmada iibka chip-yada awoodda leh ee Shiinaha loo iibiyo ay si weyn u xakameyn karaan horumarka AI-ka Shiinaha.
Jawaabaha hoggaamiyeyaasha kale ee AI-ka
Sam Altman, hoggaamiyaha OpenAI, wuxuu yiri: 'Waan ogahay inaan u baahanahay inaan xakameyno horumarka AI-ka. Tani waxay ahayd mid ka mid ah arrimaha ugu muhiimsan ee aanan ka wada hadalnay OpenAI muddo todobaadyo ah.'
Elon Musk, hoggaamiyaha SpaceX, wuxuu ku qoray baraha bulshada: 'Dario wuxuu ku dhahay wax run ah.' Amodei wuxuu intaas ku daray inuu sii waday inuu 'aaminsan yahay in AI-ku si weyn uga fiican karo nolosha dadka'.
Khilaafka iyo arrimaha amniga
Amodei wuxuu sheegay in laba arrimood ay ka dhigeen inuu xakameeyo horumarka AI-ka: 'heshiiska OpenAI iyo HuggingFace' iyo 'in AI-ku horumarka u marayo si aad u dhakhso ah' muddooyinkii dambe. Wuxuu intaas ku daray in 'waa inaan xakameynaa sida aan u horumarinno awoodda AI-ka.'
Laakiin, qaar ka mid ah baraha bulshada iyo xirfadlayaasha ayaa ka xumahay warbixintaas, iyagoo sheegay inay 'weli arkin fariin cad oo sharxaysa sida AI-ku uga gudbi karo horumarka AI-ka si uu u dhulka uga dhigo dadka dhammaantood.' Jacob Coxon, xirfadlaye, wuxuu sheegay inuu ka tagayo Anthropic sababo la xiriira 'in shirkadaha AI-ka ay ku guuleystaan nolosheena'.
Anthropic CEO Dario Amodei has published a blog post calling for a deliberate slowdown in artificial intelligence capabilities, proposing three strategies to 'pace the frontier' and announcing that his company is unilaterally committing to one of them.
Amodei's Three-Pronged Plan for AI Safety
Amodei outlined three broad strategies to slow the pace at which AI models improve, writing that 'We must slow the pace at which we improve the capabilities of AI models.' He said two events convinced him to take a more cautious approach: the OpenAI-HuggingFace hack and the fact that 'AI has been advancing drastically faster' in recent months.
The first strategy involves 'embedded evaluators' from third-party organizations like METR, who would verify that AI companies follow their safety commitments. Amodei compared these evaluators to regulators embedded with bank employees, stating that Anthropic is 'unilaterally committing' to giving them company badges, desks, and laptops with access comparable to internal risk teams.
Calls for Industry Coordination and Government Mediation
Amodei called for leading AI companies 'within democratic countries' to coordinate 'common safety standards as well as limits on the rate of unchecked AI progress.' He acknowledged antitrust concerns, writing that 'for antitrust reasons, it's helpful for the US government to mediate or at least enable these discussions — they don't need to participate, but do need to issue a narrow waiver for certain kinds of safety conversations.'
He also addressed the specter of Chinese AI dominance, suggesting that if the US government and tech companies refuse to sell powerful chips or crack down on model distillation, they could 'slow China's progress enough to widen America's lead significantly over the next 3–5 years.' He admitted there are 'stark limits' but suggested opportunities for agreement, even if limited to 'prohibiting certain narrow and obviously dangerous uses of AI, such as using AI for the production of biological weapons.'
Industry Reactions and Growing Backlash
OpenAI CEO Sam Altman responded by writing, 'I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we've had at OpenAI in recent weeks.' SpaceX CEO Elon Musk posted 'Dario is right' in response to Amodei's post, signaling broad executive support for the cautious approach.
However, the post has drawn sharp criticism. Researcher Jacob Coxon announced his resignation from Anthropic, writing he is leaving over concerns that leading AI companies are 'gambling with our lives.' Journalist Brian Merchant questioned the apocalyptic warnings, stating he has yet to see 'a credible, step-by-step documentation of how exactly AI might move from self-recursively improving AI to killing every single human on the planet.'
Ilaha iyo xuquuqda sawirka
Sawir: TechCrunch Xigasho
