Gary Marcus(@GaryMarcus)
庆祝AGI一周年快乐！
4.5Score
TL;DR · AI 摘要
AI远未达到真正通用人工智能（AGI）水平，仍频繁产生幻觉且不可信，Gary Marcus以讽刺方式提醒公众勿被误导。
核心要点
- 当前AI系统仍普遍存在幻觉问题，无法实现真正可靠的推理与决策。
- 经济学家Tyler Cowen曾错误宣称o3模型为AGI，反映部分人士对AGI标准认知偏差。
- Gary Marcus通过反讽强调：AI距离无幻觉、可信赖的AGI阶段仍有巨大差距。
结构提纲
按章节快速跳转。
作者以反语庆祝AGI一周年，实则批评当前AI系统远未达标。
所谓‘无幻觉’是讽刺，现实中大模型仍频繁生成虚假信息。
有人将o3称为AGI暴露了对通用人工智能标准的误解和过度炒作。
思维导图
用一张图看清主题之间的关系。
查看大纲文本（无障碍 / 无 JS 友好）
- 对AGI炒作的讽刺批判
- 核心立场：AI尚未达到AGI
- 存在广泛幻觉
- 缺乏可靠信任基础
- 社会认知偏差案例
- Tyler Cowen称o3为AGI
- 反映过度乐观倾向
金句 / Highlights
值得收藏与分享的关键句。
Happy One Year Anniversary, AGI! What an amazing, hallucination-free year it has been!
Glad we can totally trust you now!
I was being sarcastic.
#AGI#人工智能#幻觉问题#Gary Marcus
打开原文What an amazing, hallucination-free year it has been!*
Glad we can totally trust you now!*
*I was being sarcastic." / X
Gary Marcus on X: "Happy One Year Anniversary, AGI!* What an amazing, hallucination-free year it has been!* Glad we can totally trust you now!* *I was being sarcastic." / X
Don’t miss what’s happening
Happy One Year Anniversary, AGI!* What an amazing, hallucination-free year it has been!* Glad we can totally trust you now!* *I was being sarcastic.
Quote
@GaryMarcus
·
Jun 10, 2025
Replying to @codeslubber
tyler cowen literally said o3 was AGI.
Last edited Opens edit history 12:23 AM · May 12, 2026
·
11 15
10 17
Read 11 replies