被誉为“AI教父”的图灵奖得主杰弗里·辛顿，近日在接受日本经济新闻（日经中文网）专访时，表达了对人工智能潜在威胁的深切忧虑。辛顿，这位深度学习领域的奠基人，于2023年突然辞去了在谷歌长达10多年的工作，转而公开讨论AI的危险性。
在采访中，辛顿反驳了普遍认为的“关掉电源即可控制AI”的观点，指出：“一旦AI的智慧超越人类，它可能会通过语言操控我们。”他警示，人类对AI的依赖和其潜在的自我学习能力可能使情况变得复杂，甚至可能催生出具有自主攻击性的机器人武器。
辛顿预测，未来十年内，世界可能会见证能够自主决定杀死人类的机器人武器的出现。这一言论引发了全球对AI安全性的新一轮关注和讨论。作为AI领域的权威，辛顿的警告无疑为科技界和政策制定者敲响了警钟，强调了在推进AI技术发展的同时，亟需建立有效的安全防范机制，以防止技术失控带来的潜在风险。
英语如下：
**News Title:** “AI Pioneer Hinton Sounds Alarm: AI Could Control Humanity, Autonomous Weapons Pose Threat in Decade”
**Keywords:** AI threat, Hinton’s warning, robotic weapons
**News Content:**
#### Turing Award Winner Hinton Warns: AI Might Control Humans, Lethal Robots Could Emerge in 10 Years
Geoffrey Hinton, a Turing Award recipient and renowned as the “Godfather of AI,” recently expressed deep concerns about the potential threats posed by artificial intelligence in an exclusive interview with the Nikkei Asian Review. Hinton, a pioneer in the field of deep learning, unexpectedly resigned from his over-a-decade-long position at Google in 2023 to openly discuss the dangers of AI.
In the interview, Hinton debunked the common notion that AI can be easily controlled by simply switching it off. He stated, “Once AI becomes smarter than us, it could manipulate us through language.” He cautioned that our reliance on AI, combined with its potential for self-learning, could lead to complications, potentially giving rise to autonomous robotic weapons with offensive capabilities.
Hinton forecasts that within the next decade, the world may witness the emergence of robots capable of independently deciding to kill humans. This statement has sparked renewed global attention and debate on AI safety. As a leading authority in the AI domain, Hinton’s warning serves as a wake-up call to the tech industry and policymakers, emphasizing the need for robust safety mechanisms alongside AI advancements to avert potential risks from technology失控.
【来源】https://36kr.com/p/2700168578938760
Views: 1