請更新您的瀏覽器

您使用的瀏覽器版本較舊,已不再受支援。建議您更新瀏覽器版本,以獲得最佳使用體驗。

此為 AI 翻譯內容,經編輯審核後發布。

DeepMind 共同創辦人警告 AI 能力不得超前安全管控,FT 報導

Reuters

更新於 6小時前 • 發布於 13小時前
資料照片:這張攝於 2025 年 1 月 27 日的示意圖中,可見寫著「AI 人工智慧」的訊息、一個鍵盤和機器人雙手。路透社/Dado Ruvic/示意圖/資料照片FILE PHOTO: A message reading "AI artificial intelligence", a keyboard, and robot hands are seen in this illustration taken January 27, 2025. REUTERS/Dado Ruvic/Illustration/File Photo

路透 9 月 16 日電——英國《金融時報》週三報導,Google DeepMind 共同創辦人夏恩・萊格(Shane Legg)警告,快速進展的 AI 不得超前安全管控;他同時成立一間研究院,探索通用人工智慧(AGI)的部署。Sept 16 (Reuters) - Google DeepMind co-founder Shane Legg warned that rapidly advancing AI must not outpace safety controls, as he launched an institute to explore the deployment of artificial general intelligence (AGI), the Financial Times reported on Wednesday.

AI 公司對遞迴式自我改進提出警告之際,研究人員指出,能力日益增強的 AI 代理人已能撰寫軟體、執行自主任務,並協助開發 AI 系統。這引發外界憂慮,擔心這項技術的進展速度可能超過監測與控制它的防護措施。AI companies' warnings about recursive self-improvement come as researchers report that increasingly capable AI agents can write software, perform autonomous tasks and assist in developing AI systems. This raises concerns that the technology could advance faster than safeguards to monitor and control it.

• 報導指出,萊格表示,Anthropic 執行長達里歐・阿莫迪(Dario Amodei)呼籲放慢但不要暫停發布前沿模型,這在方向上「很有意思」,也「值得考慮」。• Anthropic CEO Dario Amodei's call to slow but not pause the release of frontier models was "interesting directionally" and "worth considering," Legg said, according to the report.

• 《金融時報》報導,儘管輝達(Nvidia)和 OpenAI 高層近來聲稱 AGI 已經實現,但萊格表示,現在宣布 AGI 已達成仍為時過早。• He said it was premature to declare that AGI had been achieved, despite recent claims by executives at Nvidia and OpenAI, the FT reported.

• 報導稱,萊格仍然「認同」自己的預測,也就是到 2028 年實現「最低限度」AGI 的機率為 50%。• Legg remains "comfortable" with his forecast of a 50% chance of achieving "minimal" AGI by 2028, the report said.

• 報導補充說,DeepMind 研究院(DeepMind Institute)旨在擴大公眾對 AGI 的討論;Google DeepMind 將 AGI 定義為具有人腦認知能力的系統。• The DeepMind Institute aims to broaden public discussion of AGI, which Google DeepMind defines as a system with the cognitive capabilities of the human brain, the report added.

• 《金融時報》指出,該研究院將涵蓋科學、教育、社會、政策、哲學和人類繁榮等議題。• The institute will cover topics including science, education, society, policy, philosophy and human flourishing, according to the FT.

• 報導稱,該研究院首批三篇文章探討 AI 安全、經濟政策,以及「新烏托邦主義」的原則。• Its first three essays address AI safety, economic policy and principles for a "new utopianism," the report said.

(班加羅爾 Anzar Mehraj 報導;Shreya Biswas 編輯)(Reporting by Anzar Mehraj in Bengaluru; Editing by Shreya Biswas)

更多國際相關文章

01

賣一台電梯能穩賺70年 華爾街重新發現隱形金礦

TVBS
02

旅日台籍醫師「背景超狂」涉性侵遭逮捕 患者紛紛大喊:他是親切的好人!

鏡報
03

震碎三觀!母子吵架控家暴意外揭亂倫案 媽媽:和兒子上床100次了

鏡報
04

吃外送咬到「寵物晶片」還掃出編號!一查竟是這動物

民視新聞網
05

律師逛夜市「遭磚塊爆頭亡」!兇手6天「狂丟水桶、可樂」只為判死刑:砸死誰算誰

鏡報
06

黛安娜逝世將滿30週年!親弟爆料當年查爾斯冷血對話:我們很快會忘了她

太報
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...