Anthropic創辦人示警
Artificial intelligence is developing at an astonishing speed, and one of its leading creators is warning the world to be careful.
人工智慧（AI）正以驚人的速度發展，而其中一位領軍人物正警告世人務必謹慎。
astonishing(adj.)驚人的、令人震驚的leading(adj.)領先的;首屈一指的;重要的prominent(adj.)著名的;傑出的;重要的major(adj.)主要的
Jack Clark, co-founder of Anthropic, recently told the BBC that the AI industry has a powerful “gas pedal” but no “brake pedal.”
美國AI新創公司Anthropic的共同創辦人傑克·克拉克近日向BBC表示，AI產業雖擁有強大的「油門」，卻缺乏「煞車」。
gas pedal油門踏板pedal to the metal全力加速;全力衝刺;加足馬力
He believes AI developers need the ability to slow down or even pause development when necessary.
他認為，AI開發者必須具備在必要時能減速甚至暫停開發的能力。
slow down減速speed up加速=acceleratewhen necessary在需要時=when needed
This concern comes as AI systems like Anthropic’s Claude are already writing most of their own code — more than 80% today and possibly 100% within two years.
在克拉克提出此擔憂之際，Anthropic的Claude等AI系統已能自行編寫大部分程式碼——目前比例已超過80%，且可能在兩年內達到100%。
The main concern is that AI might soon improve itself without much human help, a process called “recursive self-improvement.”
主要擔憂在於，人工智慧可能很快就能在幾乎不需要人類協助的情況下自我提升，這個過程被稱為「遞迴式自我改進」。
assistance(n.)協助=aidrecursive self-improvement遞迴式自我改進recursive(adj.)遞迴的;反覆呼叫自身的;循環式的
This could bring huge benefits, such as solving difficult problems in medicine or climate change.
這雖能帶來巨大好處，例如解決醫學或氣候變遷等棘手問題。
thorny(adj.)棘手的;複雜難解的tricky(adj.)棘手的;困難的knotty(adj.)難以解決的;棘手的
However, it also raises serious risks, including loss of control and major job losses.
然而，也伴隨著嚴重的風險，包括人類失去控制權以及大規模失業。
raise a concern引起擔憂
Many companies have already used AI to replace workers, especially in software engineering.
許多企業已開始運用AI取代人力，特別是在軟體工程領域。
especially(adv.)尤其是;特別是=particularly
Clark points out that while AI may struggle with true creativity for now, its rapid growth could still disrupt economies and societies worldwide.
克拉克指出，儘管人工智慧目前在真正的創造力方面仍有不足，但其快速發展仍可能對全球經濟與社會造成顛覆性衝擊。
struggle with...在某方面遇到難題disrupt(v.)擾亂、顛覆
To manage these challenges, Clark and Anthropic suggest governments create clear regulations, similar to how society once controlled the powerful oil industry.
為應對這些挑戰，克拉克與Anthropic建議各國政府制定明確的法規，類似於社會過去管控強大的石油產業的方式。
regulation(n.)法規、監管措施regulatory(adj.)監管的regulatory body監管機構similar to...類似於...
They also call for coordinated pauses among major AI companies so that safety research can catch up.
他們還呼籲主要AI公司協調暫停研發，讓安全研究能跟上腳步。
call for呼籲;公開要求coordinate(v.)協調;使相配合coordinated(adj.)協調的catch up追上、趕上
Without global agreement — especially between the US and China — competitive pressure might make slowdowns impossible.
若缺乏全球共識——特別是美中之間——競爭壓力可能會使研發放緩變得難以實現。
consensus(n.)共識competitive(adj.)有競爭力的、競爭激烈的
As AI becomes increasingly powerful, what should we do?
隨著AI變得越來越強大，我們該怎麼應對？
increasingly(adv.)越來越=more and moreprogressively(adv.)逐漸地
Clark advises young people to develop hobbies, read widely, and study liberal arts.
克拉克建議年輕人培養興趣、廣泛閱讀，並接受人文教育。
advise sb. to do sth.建議某人做某事extensively(adv.)廣泛地liberal(adj.)人文的
Creative and curious minds will still have an advantage.
富有創造力與好奇心的人仍將擁有優勢。
mind(n.)聰明人;有才智的人competitive edge競爭優勢
This moment invites all of us to think seriously: How can we enjoy AI’s benefits while keeping human values and control at the center?
此刻，我們所有人都應認真思考：如何在享受AI帶來的好處同時，牢牢守住人類價值與控制權？
keep...at the center讓……保持在核心、中心位置
The conversation is just beginning.
這場對話才正要開始。