Meta 首席科学家、图灵奖得主 Yann LeCun 近日在接受知名科技博主 Lex Fridman 的深度访谈时明确表示，他认为人工智能（AI）毁灭人类的可能性为零。这场长达三个小时的对话涵盖了广泛的 AI 主题，包括大语言模型（LLM）的局限性、大模型的幻觉、人工智能通用性（AGI）的挑战，以及对普遍存在的 AI 末日论的批判。
LeCun 在访谈中指出，虽然 AI 技术在许多领域展现出惊人的进步，但目前的系统仍然存在显著的局限性，不足以构成对人类生存的威胁。他强调，公众和媒体对于 AI 的理解往往存在误解，即所谓的“大模型幻觉”，即认为这些模型具有远超其实力的智能和自主性。
在讨论 AGI 时，LeCun 指出，尽管这是一个长期的研究目标，但实现这一目标将面临诸多科学和技术上的挑战。他认为，目前的 AI 系统远未达到能够独立思考和决策的程度，因此，对于 AI 末日的担忧更多是科幻想象而非现实风险。
这场深入的访谈为公众提供了一个理解 AI 的新窗口，帮助人们理性看待 AI 的发展，认识到其潜力的同时，也要清醒地看到其限制。LeCun 的观点无疑为 AI 技术的发展注入了一剂理智的强心针，提醒人们在追求科技进步的同时，不应忽视人类的控制和伦理考量。
英语如下：
**News Title:** “Turing Award Winner Yann LeCun: AI’s Chance of Doomsday is Zero, In-Depth Interview Offers New Perspectives on AI”
**Keywords:** Yann LeCun, AI Safety, Chief Scientist at Meta
**News Content:** In a recent extensive interview with renowned tech blogger Lex Fridman, Yann LeCun, Chief Scientist at Meta and recipient of the Turing Award, categorically stated that he believes the probability of artificial intelligence (AI) dooming humanity is zero. The three-hour conversation delved into a broad range of AI topics, including the limitations of large language models (LLMs), illusions surrounding large models, the challenges of artificial general intelligence (AGI), and a critique of prevalent AI doomsday scenarios.
During the discussion, LeCun pointed out that while AI has shown remarkable progress in numerous fields, current systems still have significant limitations that do not pose an existential threat to humans. He emphasized that there is often a misunderstanding among the public and media, referred to as the “large model illusion,” which overestimates these models’ intelligence and autonomy.
When discussing AGI, LeCun noted that while it remains a long-term research goal, achieving it will entail numerous scientific and technological hurdles. He argued that current AI systems are far from being capable of independent thought and decision-making, thus, concerns about an AI doomsday are more rooted in science fiction than real-world risks.
This in-depth interview offers a new perspective for the public to understand AI, encouraging a balanced view of its advancements and limitations. LeCun’s insights serve as a rational boost for AI development, reminding us to pursue technological progress while not overlooking human control and ethical considerations.
【来源】https://mp.weixin.qq.com/s/ny6awp7BMvnRCZbn4lLTsA
Views: 7