Meta 的首席科学家、图灵奖获得者 Yann LeCun 近日在一个深入的科技访谈中向知名博主 Lex Fridman 表达了他对人工智能（AI）未来的看法。在长达近三个小时的对话中，LeCun 公开反驳了关于 AI 毁灭人类的担忧，他认为这种可能性为零。这场内容丰富的访谈涵盖了从限制性语言模型（LLM）的局限性，到大型模型的误解，再到通用人工智能（AGI）所面临的挑战等广泛议题。
LeCun 在采访中指出，尽管 AI 技术正在迅速发展，但目前的系统仍然存在显著的局限性，远未达到能够威胁人类生存的程度。他强调，公众和部分专家对 AI 的理解往往被“大模型幻觉”所误导，即过高估计了这些系统的能力和自主性。他认为，AI 的发展需要克服重重障碍，包括理解复杂环境、具备常识推理以及自我意识的构建，这些在当前的技术水平下都是巨大的挑战。
此外，LeCun 的观点也对 AI 末日论进行了批判，他认为过度担忧 AI 的潜在风险可能会阻碍科技创新和有益应用的推进。他提倡在保持警惕的同时，应当更加积极地探索 AI 的潜力，以解决现实世界的问题。
这场访谈为公众提供了一个深入理解 AI 的窗口，同时也提醒了科技界在追求技术进步的同时，需要理性看待 AI 的能力与限制，以确保其发展对人类社会带来的是积极影响。来源：AppSo。
英语如下：
**News Title:** “Turing Award Winner Yann LeCun: AI poses zero risk to humanity, in-depth interview sheds new light on AI perspectives”
**Keywords:** Yann LeCun, AI safety, Chief Scientist at Meta
**News Content:** In a recent extensive technology interview with renowned podcaster Lex Fridman, Yann LeCun, Chief Scientist at Meta and recipient of the Turing Award, shared his insights on the future of artificial intelligence (AI). During the nearly three-hour conversation, LeCun debunked concerns about AI’s potential to毁灭人类, stating that such a probability is zero. The wide-ranging discussion touched upon topics from the limitations of large language models (LLMs) to misconceptions about large models and the challenges facing artificial general intelligence (AGI).
LeCun pointed out during the interview that while AI technology is advancing rapidly, current systems still have significant limitations, falling far short of posing an existential threat to humans. He cautioned that both the public and some experts often fall prey to the “Big Model Illusion,” overestimating these systems’ capabilities and autonomy. He believes that overcoming hurdles like understanding complex environments, common sense reasoning, and self-awareness construction – all formidable challenges at present technological levels – is crucial for AI’s development.
Moreover, LeCun criticized doomsday scenarios surrounding AI, arguing that excessive worry about potential risks could stifle innovation and beneficial applications. Advocating for a cautious yet proactive approach to exploring AI’s potential, he emphasized the need to address real-world problems while acknowledging AI’s capabilities and constraints.
This interview offers the public a deeper understanding of AI and serves as a reminder to the tech community to approach technological advancements with a rational view of AI’s abilities and limitations, ensuring its development has a positive impact on society. _Source: AppSo._
【来源】https://mp.weixin.qq.com/s/ny6awp7BMvnRCZbn4lLTsA
Views: 1