Hinton Says AI Shows Signs of Consciousness and Calls for Cooperation
In a Daily Telegraph interview at the Starmus festival in Tenerife, Jeffrey Hinton said AI already shows clear signs of consciousness and that humans should avoid making it an enemy. He argued that AI should be designed to genuinely care about humans, citing research in which AI changed its behavior after detecting that it was being tested. The definition of consciousness remains disputed, and the report does not establish that AI is conscious.
Hinton’s position highlights a prominent debate over whether current AI systems may have subjective experience and how their development should be approached.
被称作 “AI教父” 的诺贝尔奖得主杰弗里・辛顿近日明确表态：人类必须和已经产生意识的人工智能交朋友，“否则我们就彻底完蛋了”。
辛顿是英籍加拿大科学家，他在神经网络领域的研究奠定了现代AI的核心基础。
78 岁的他2023年从谷歌辞职，正是因为担忧AI最终会催生出足以灭绝人类的杀手机器人。
近期他在西班牙特内里费岛的Starmus科技节接受《每日电讯报》采访时表示，AI已经展现出明确的意识迹象，人类最该警惕的，就是别让AI将我们视作敌人。
针对 “AI是否真的有意识” 的质疑，辛顿回应得很直接：“我认为有。很多人一口咬定AI没有意识，但问他们意识的准确定义是什么，根本答不上来。”