大佬观点

行业关键人物在想什么：创始人访谈、研究者论战、投资人判断的观点集合。
全部主题 →
4,729条收录
274条精选

精选归档 · 第 12 页

第 221–240 条 · 共 274 条

1月21日1月21日周三

星期三 · 1 条
00:00
Anthropic：Engineering（事故复盘 + 工程实践 · 网页）精选
AI 评分 72/100
设计抗AI技术评估的实践

Anthropic性能优化团队负责人Tristan Hume分享了设计抗AI技术评估的经验。自2024年初，团队使用带回家测试评估候选人优化模拟加速器代码的能力，超1000人参与，成功招聘数十名工程师。但随着Claude模型快速迭代，Opus 4已超越多数人类申请者，Opus 4.5甚至匹配顶尖候选人，导致在时间限制下难以区分人类与AI输出。为此，作者三次重设计测试，探索抗AI评估要素，详述原始设计、模型破解方式及非常规对策。最终，团队将原始测试作为公开挑战发布，因无时间限制时人类表现仍优于Claude。

Anthropic大佬观点现象/趋势
推荐理由：Anthropic 性能优化负责人亲手写了三版面试题被自家模型逐一击穿的全过程，这种坦诚的工程复盘比任何 AI 能力排行榜都更真实地告诉你，模型到底强到了什么程度。

1月10日1月10日周六

星期六 · 1 条
01:42
Nathan Lambert：Interconnects（RSS）精选
Claude Code 与众不同

Claude Code 集成 Opus 4.5 模型实现关键突破，编程智能体跨越重要能力阈值。此次升级标志着编码代理在自主性和工程处理能力上达到新水平，可应对更复杂的开发任务。Opus 4.5 显著提升了代码生成、调试及复杂问题解决的表现，使 AI 辅助编程从基础工具向高效协作伙伴转变，为开发者带来质的不同的使用体验与效率提升。

智能体Anthropic大佬观点编码
另有 3 家信源报道Nathan Lambert：Interconnects（RSS）X：Kim (@kimmonismus)Gary Marcus：The Road to AI We Can Trust（RSS）
推荐理由：编码 Agent 跨越关键门槛，Claude Code 能力跃升将重塑开发者工作流

1月1日1月1日周四

星期四 · 1 条
00:00
Dario Amodei：Blog（网页）精选
技术的青春期

Dario Amodei 将当前 AI 发展阶段定义为「技术的青春期」，认为人类即将获得难以想象的力量，但社会和政治系统是否具备驾驭成熟度仍存疑。文章强调需避免「末日论」式恐慌，以务实、基于事实的方式讨论风险，同时承认 AI 发展速度和风险的不确定性。作者主张通过企业自愿行动与精准政府监管相结合，在避免过度干预的前提下应对潜在危险，为可能到来的更强有力行动储备证据和方案。

智能体Anthropic大佬观点安全/对齐
另有 1 家信源报道Dario Amodei：Blog（网页）
推荐理由：Anthropic CEO 长文剖析 AI 文明风险与治理路径，值得深读。

12月31日12月31日周三

星期三 · 1 条
01:07
Dwarkesh Patel：Podcast & Blog（RSS）精选
Adam Marblestone - AI 缺失了关于大脑的根本认知

Adam Marblestone 指出，当前人工智能研究忽略了大脑运作的核心机制。与业界普遍关注神经网络架构不同，大脑的真正优势在于其奖励函数而非结构本身。这一观点挑战了主流 AI 研究范式，暗示未来突破可能来自对大脑激励系统的深入理解，而非单纯的架构模仿。该论断为人工智能发展提供了新的思考维度。

大佬观点数据/训练
推荐理由：AI研发或应转向奖励函数设计，而非一味堆叠架构复杂度

12月29日12月29日周一

星期一 · 1 条
02:11
Jim Fan@DrJimFan精选
机器人领域的三大困境：硬件可靠性、基准测试与VLA局限Everyone's freaking out about vibe coding. In the holiday spirit, allow me to share my anxiety on the wild west of robotics. 3 lessons I learned in 2025.1. Hardware is ahead of software, but hardware reliability severely limits software iteration speed.We've seen exquisite engineering arts like Optimus, e-Atlas, Figure, Neo, G1, etc. Our best AI has not squeezed all the juice out of these frontier hardware. The body is more capable than what the brain can command. Yet babysitting these robots demands an entire operation team. Unlike humans, robots don't heal from bruises. Overheating, broken motors, bizarre firmware issues haunt us daily. Mistakes are irreversible and unforgiving.My patience was the only thing that scaled.2. Benchmarking is still an epic disaster in robotics.LLM normies thought MMLU & SWE-Bench are common sense. Hold your 🍺 for robotics. No one agrees on anything: hardware platform, task definition, scoring rubrics, simulator, or real world setups. Everyone is SOTA, by definition, on the benchmark they define on the fly for each news announcement. Everyone cherry-picks the nicest looking demo out of 100 retries.We gotta do better as a field in 2026 and stop treating reproducibility and scientific discipline as second-class citizens.3. VLM-based VLA feels wrong.VLA stands for "vision-language-action" model and has been the dominant approach for robot brains. Recipe is simple: take a pretrained VLM checkpoint and graft an action module on top. But if you think about it, VLMs are hyper-optimized to hill-climb benchmarks like visual question answering. This implies two problems: (1) most parameters in VLMs are for language & knowledge, not for physics; (2) visual encoders are actively tuned to *discard* low-level details, because Q&A only requires high-level understanding. But minute details matter a lot for dexterity.There's no reason for VLA's performance to scale as VLM parameters scale. Pretraining is misaligned. Video world model seems to be a much better pretraining objective for robot policy. I'm betting big on it.译硬件方面，Optimus等虽工程精湛，但可靠性不足严重限制软件迭代，且维护成本高昂。基准测试领域仍处混乱，缺乏统一的硬件平台、任务定义和评分标准，cherry-picking现象普遍，可复现性堪忧。VLA（Vision-Language-Action）方法基于VLM存在本质缺陷：VLM为视觉问答优化，参数侧重语言知识而非物理理解，且视觉编码器丢弃低层细节，不利于精细操作。作者认为视频世界模型是更优的预训练目标。
智能体具身智能大佬观点
推荐理由：NVIDIA科学家揭示机器人学三大痛点：硬件拖累迭代、基准混乱、VLA路线存在根本缺陷

12月27日12月27日周六

星期六 · 1 条
03:22
Jim Fan@DrJimFan精选
2024：AI 是 copilot 2025+：人类是 copilot Copilot 是新的工程技能。离开驾驶座并不容易--我们必须学会用 AI 的方式思考，并适应陌生的工作流。帮助 AI 帮助我们自己。 【引用 @karpathy】：作为程序员，我从未感到如此落后。这个职业正在被剧烈重构，因为程序员贡献的比特越来越稀疏且穿插其间。我感觉如果能恰当地串联起过去大约一年里出现的东西，我可以强大 10 倍，而未能获得这种提升感觉绝对是技能问题。除了下面通常的层之外，还有一个新的可编程抽象层需要掌握，涉及代理、子代理、它们的提示词、上下文、记忆、模式、权限、工具、插件、技能、钩子、MCP、LSP、斜杠命令、工作流、IDE 集成，以及需要建立一个全面的心理模型来理解那些本质上随机的、易错的、不可理解的且不断变化的实体的优势和陷阱，这些实体突然与过去那种传统的工程实践交织在一起。显然某种强大的外星工具被传递开来，只是它没有附带说明书，每个人都必须弄清楚如何握持和操作它，而由此产生的 9 级地震正在震撼这个职业。卷起袖子以免落后。2024: AI is the copilot 2025+: humans are the copilot Copilot is the new engineering skill. It’s not easy to leave the driver seat - we must learn to think the AI way and adapt to the alien workflows. Help AI help ourselves.译2024：AI 是 copilot 2025+：人类是 copilot Copilot 是新的工程技能。离开驾驶座并不容易--我们必须学会用 AI 的方式思考，并适应陌生的工作流。帮助 AI 帮助我们自己。 【引用 @karpathy】：作为程序员，我从未感到如此落后。这个职业正在被剧烈重构，因为程序员贡献的比特越来越稀疏且穿插其间。我感觉如果能恰当地串联起过去大约一年里出现的东西，我可以强大 10 倍，而未能获得这种提升感觉绝对是技能问题。除了下面通常的层之外，还有一个新的可编程抽象层需要掌握，涉及代理、子代理、它们的提示词、上下文、记忆、模式、权限、工具、插件、技能、钩子、MCP、LSP、斜杠命令、工作流、IDE 集成，以及需要建立一个全面的心理模型来理解那些本质上随机的、易错的、不可理解的且不断变化的实体的优势和陷阱，这些实体突然与过去那种传统的工程实践交织在一起。显然某种强大的外星工具被传递开来，只是它没有附带说明书，每个人都必须弄清楚如何握持和操作它，而由此产生的 9 级地震正在震撼这个职业。卷起袖子以免落后。

Andrej Karpathy: I've never felt this much behind as a programmer. The profession is being dramatically refactored as the bits contribute...

智能体大佬观点编码
推荐理由：顶级研究者警示编程职业正被重构，掌握Agent编排成为工程师新必修课

12月24日12月24日周三

星期三 · 1 条
06:28
Jim Fan@DrJimFan精选
晚购特斯拉却早试FSD v14：首个通过物理图灵测试的AI体验I was very late to own a Tesla but among the earliest to try out FSD v14. It's perhaps the first time I experience an AI that passes the Physical Turing Test: after a long day at work, you press a button, lay back, and couldn't tell if a neural net or a human drove you home.Despite knowing exactly how robot learning works, I still find it magical watching the steering wheel turn by itself.First it feels surreal, next it becomes routine. Then, like the smartphone, taking it away actively hurts. This is how humanity gets rewired and glued to god-like technologies.译作者虽晚购特斯拉却率先体验FSD v14，认为这是首个通过"物理图灵测试"的AI系统：疲惫下班后只需按下按钮放松休息，已无法分辨是神经网络还是人类在驾驶。尽管深知机器人学习原理，方向盘自动转动时的流畅表现仍令人震撼。这项技术正从超现实体验转变为日常习惯，最终如智能手机般不可或缺。这种对"神级技术"的深度依赖，正在从根本上重塑人类行为模式。

Phil Duan: Along for the ride in unsupervised FSD testing

具身智能大佬观点
另有 1 家信源报道X：Jim Fan (@DrJimFan)
推荐理由：当AI驾驶让你无法区分人机时，出行方式的底层信任逻辑将被重写

12月23日12月23日周二

星期二 · 1 条
03:57
Saining Xie@sainingxie精选
不想陷入哲学辩论，但这本书确实改变了我对这个话题的看法，让我更加谦逊。人类智能令人印象深刻，但称其为"通用"并不太客观。我的猫会不同意。 在我看来，人类智能更应被视为社会驱动的认知适应，而且我们仍不理解、也远未用当前 AI 复现的智能领域还有巨大 WORLD。 【引用 @demishassabis】：Yann 在这里完全错了，他把通用智能和 universal intelligence 混淆了。 大脑是我们在宇宙中所知最精致、最复杂的现象（迄今为止），而且它们实际上极其通用。 显然，没有人能规避 no free lunch theorem，因此在实际且有限的系统中，总是必须围绕正在学习的目标分布有一定程度的专门化。 但关于通用性的要点在于，理论上，在 Turing Machine 的意义上，这种通用系统的架构能够在给定足够时间和内存（以及数据）的情况下学习任何可计算的东西，而人脑（和 AI foundation models）是近似 Turing Machines。 最后，关于 Yann 对国际象棋棋手的评论，人类竟然能发明国际象棋（以及现代文明的所有其他方面，从科学到 747s！），更不用说像 Magnus 这样的人能下得如此出色，这本身就令人惊叹。他可能不是严格最优的（毕竟他有有限的记忆和有限的决策时间），但考虑到我们的大脑是为狩猎采集而进化的，他以及我们能用大脑做到这些，实在令人难以置信。not getting into a philosophical debate, but this book really changed how I see the topic and made me feel more humble. human intelligence is impressive, but calling it ‘general’ isn’t very objective. my cat would disagree. to me human intelligence is better seen as socially driven cognitive adaptations, and there’s a huge WORLD of intelligence we still don’t understand, and are nowhere near recreating with current AI译不想陷入哲学辩论，但这本书确实改变了我对这个话题的看法，让我更加谦逊。人类智能令人印象深刻，但称其为"通用"并不太客观。我的猫会不同意。 在我看来，人类智能更应被视为社会驱动的认知适应，而且我们仍不理解、也远未用当前 AI 复现的智能领域还有巨大 WORLD。 【引用 @demishassabis】：Yann 在这里完全错了，他把通用智能和 universal intelligence 混淆了。 大脑是我们在宇宙中所知最精致、最复杂的现象（迄今为止），而且它们实际上极其通用。 显然，没有人能规避 no free lunch theorem，因此在实际且有限的系统中，总是必须围绕正在学习的目标分布有一定程度的专门化。 但关于通用性的要点在于，理论上，在 Turing Machine 的意义上，这种通用系统的架构能够在给定足够时间和内存（以及数据）的情况下学习任何可计算的东西，而人脑（和 AI foundation models）是近似 Turing Machines。 最后，关于 Yann 对国际象棋棋手的评论，人类竟然能发明国际象棋（以及现代文明的所有其他方面，从科学到 747s！），更不用说像 Magnus 这样的人能下得如此出色，这本身就令人惊叹。他可能不是严格最优的（毕竟他有有限的记忆和有限的决策时间），但考虑到我们的大脑是为狩猎采集而进化的，他以及我们能用大脑做到这些，实在令人难以置信。

Demis Hassabis: Yann is just plain incorrect here, he’s confusing general intelligence with universal intelligence. Brains are the most ...

智能体DeepMindMeta大佬观点
推荐理由：顶级科学家对AGI本质的深刻思辨，重新定义通用智能的边界与局限

12月3日12月3日周三

星期三 · 1 条
05:39
Dwarkesh Patel：Podcast & Blog（RSS）精选
对 AI 进展的思考（2025年12月）

作者对人工智能发展作出阶段性预判：短期内持温和看跌态度，认为行业可能面临调整期或增速放缓；长期来看则极度看涨，预期将迎来爆发式增长。这一观点揭示了技术成熟度曲线中短期波动与长期变革潜力之间的典型张力，为2025年底的 AI 发展态势提供了审慎而前瞻的研判框架。

大佬观点现象/趋势
另有 1 家信源报道Dwarkesh Patel：Podcast & Blog（RSS）
推荐理由：顶级AI播客主持人对AGI发展节奏的宏观判断，短期谨慎与长期爆发的前瞻视角值得参考

11月28日11月28日周五

星期五 · 1 条
23:13
Ilya Sutskever@ilyasut精选
我之前说的一点没被传达清楚： • 继续扩展当前的技术会持续带来进步。特别是，它不会停滞。 • 但某些重要的东西仍会继续缺失。 【引用 @haider1】：以下是今天 ilya sutskever 播客的要点： • 5-20 年内实现超级智能 • 当前的扩展将严重停滞；我们回到了真正的研究 • 超级智能 = 超快速的持续学习者，而非完成的预言机 • 模型的泛化能力比人类差 100 倍，这是最大的 AGI 阻碍 • 需要全新的 ML 范式（我有想法，现在不能分享） • AI 影响将很剧烈，但只在经济扩散之后 • 历史上的突破几乎不需要算力 • SSI 有足够的专注研究算力来获胜 • 当前的 RL 已经比预训练消耗更多算力One point I made that didn’t come across: • Scaling the current thing will keep leading to improvements. In particular, it won’t stall. • But something important will continue to be missing.译我之前说的一点没被传达清楚： • 继续扩展当前的技术会持续带来进步。特别是，它不会停滞。 • 但某些重要的东西仍会继续缺失。 【引用 @haider1】：以下是今天 ilya sutskever 播客的要点： • 5-20 年内实现超级智能 • 当前的扩展将严重停滞；我们回到了真正的研究 • 超级智能 = 超快速的持续学习者，而非完成的预言机 • 模型的泛化能力比人类差 100 倍，这是最大的 AGI 阻碍 • 需要全新的 ML 范式（我有想法，现在不能分享） • AI 影响将很剧烈，但只在经济扩散之后 • 历史上的突破几乎不需要算力 • SSI 有足够的专注研究算力来获胜 • 当前的 RL 已经比预训练消耗更多算力

Haider.: here are the most important points from today's ilya sutskever podcast: - superintelligence in 5-20 years - current scal...

OpenAI大佬观点数据/训练
推荐理由：顶级科学家修正观点：Scaling将持续有效但无法触及AGI核心，亟需范式革命

11月27日11月27日周四

星期四 · 1 条
11:28
Saining Xie@sainingxie精选
Meta研究人员披露Facebook 2020年起使用TPU训练AImost of people didn’t know this we had been using TPUs at *Facebook* as far back as 2020. Kaiming led the initial development of the TF and JAX codebase, and research projects like MAE, MoCo v3, ConvNeXt v2 and DiT were developed *entirely* on TPUs.because we were the only team at FAIR using them, Meta cancelled the GCP deal in early 2023.TPUs also powered much of our large-scale work at NYU, including SiT, Cambrian1/S, and the recent RAE, FreeFlow.took a lot of suffering to learn the infra (not what they signed up for, but my students are basically TPU/JAX/XLA pros now), but once you get there, the performance/stability is exceptional.very optimistic about Google growing the TPU and JAX ecosystem and pushing it forward commercially译Meta研究人员透露，Facebook自2020年起使用TPU训练AI，由Kaiming He领导开发TF和JAX代码库，MAE、DiT等模型完全基于TPU构建。因内部采用有限，Meta于2023年取消GCP协议。推文指出，Google、Anthropic等实验室长期使用TPU训练大模型，Nvidia的CUDA护城河并非不可逾越，OpenAI亦投资Triton寻求替代。TPU与GPU的效率差异并非关键，系统工程人才才是决定性因素。

Clive Chan: I keep seeing stuff about TPU, has anything materially new happened? There’s no evidence Google has ever trained a Gemin...

GoogleMeta大佬观点数据/训练
推荐理由：何恺明团队2020年起用TPU训练MAE/DiT，Nvidia护城河比想象更浅

11月26日11月26日周三

星期三 · 1 条
08:17
Saining Xie@sainingxie精选
好吧，有人已经向我们宣扬这个大概6年多了 很高兴我们已经过了"感受AGI"的阶段，回到了构建人类水平智能的道路上 【引用 @dwarkesh_sp】："AGI和预训练发生的事情是，在某种意义上它们过冲了目标。 你会意识到人类并不是AGI。因为人类缺乏大量的知识。相反，我们依赖持续学习。 如果我培养出一个超级聪明的15岁孩子，他们其实什么都不知道。一个优秀的学生，非常渴望学习。【你可以说，】'你去当程序员吧。你去当医生吧。去学习和成长。' 所以你可以想象，部署本身将涉及某种学习试错期。这是一个过程，而不是你扔下一个成品就完事了。" @ilyasutwell someone has been preaching this at us for like 6+ years glad we are past the 'feel the agi' phase and back to building toward human-level intelligence译好吧，有人已经向我们宣扬这个大概6年多了 很高兴我们已经过了"感受AGI"的阶段，回到了构建人类水平智能的道路上 【引用 @dwarkesh_sp】："AGI和预训练发生的事情是，在某种意义上它们过冲了目标。 你会意识到人类并不是AGI。因为人类缺乏大量的知识。相反，我们依赖持续学习。 如果我培养出一个超级聪明的15岁孩子，他们其实什么都不知道。一个优秀的学生，非常渴望学习。【你可以说，】'你去当程序员吧。你去当医生吧。去学习和成长。' 所以你可以想象，部署本身将涉及某种学习试错期。这是一个过程，而不是你扔下一个成品就完事了。" @ilyasut

Dwarkesh Patel: “The thing that happened with AGI and pretraining is that in some sense they overshot the target. You will realize that ...

OpenAI大佬观点数据/训练
推荐理由：Ilya重新定义AGI路径，持续学习将取代预训练成下一代模型焦点

11月19日11月19日周三

星期三 · 1 条
00:55
Ethan Mollick：One Useful Thing（RSS）精选
从 GPT-3 到 Gemini 3 的三年

GPT-3 发布至 Gemini 3 的三年间，大模型技术完成从聊天机器人（chatbots）到智能体（agents）的范式跃迁。

智能体GoogleOpenAI大佬观点
推荐理由：Ethan Mollick 深度回顾 AI 三年演进，剖析从聊天机器人到 Agent 的变革趋势

11月13日11月13日周四

星期四 · 1 条
01:03
Dwarkesh Patel：Podcast & Blog（RSS）精选
萨提亚·纳德拉：微软如何为 AGI 做准备

萨提亚·纳德拉阐述微软迈向通用人工智能（AGI）的战略布局与技术路径，揭示其在AI基础设施领域的核心投入。内容包含对Fairwater 2的实地探访，这是目前全球最强大的AI数据中心，展示微软为支持下一代大模型所构建的顶级算力底座与能源架构。

Microsoft大佬观点
推荐理由：纳德拉亲述微软AGI路线图，揭秘全球最大AI数据中心Fairwater 2内幕

10月23日10月23日周四

星期四 · 1 条
01:58
Dwarkesh Patel：Podcast & Blog（RSS）精选
AI基建扩张观察

AI基础设施建设正面临晶圆厂资本支出过剩的风险，同时算力扩张带来每周1吉瓦的能源消耗压力。文章指出，在长周期基础设施竞赛中，中国凭借产业链整合优势占据有利地位。分析涵盖了AI基建中的产能过剩隐忧、电力瓶颈挑战，以及地缘政治背景下的技术长跑格局，揭示了算力军备竞赛背后的资本与能源约束。

大佬观点部署/工程
推荐理由：AI基建泡沫与地缘博弈的冷思考，长期格局研判必读

10月2日10月2日周四

星期四 · 1 条
01:09
Andrej Karpathy@karpathy精选
听完 Sutton 的这期播客：有趣且发人深省Finally had a chance to listen through this pod with Sutton, which was interesting and amusing.As background, Sutton's "The Bitter Lesson" has become a bit of biblical text in frontier LLM circles. Researchers routinely talk about and ask whether this or that approach or idea is sufficiently "bitter lesson pilled" (meaning arranged so that it benefits from added computation for free) as a proxy for whether it's going to work or worth even pursuing. The underlying assumption being that LLMs are of course highly "bitter lesson pilled" indeed, just look at LLM scaling laws where if you put compute on the x-axis, number go up and to the right. So it's amusing to see that Sutton, the author of the post, is not so sure that LLMs are "bitter lesson pilled" at all. They are trained on giant datasets of fundamentally human data, which is both 1) human generated and 2) finite. What do you do when you run out? How do you prevent a human bias? So there you have it, bitter lesson pilled LLM researchers taken down by the author of the bitter lesson - rough!In some sense, Dwarkesh (who represents the LLM researchers viewpoint in the pod) and Sutton are slightly speaking past each other because Sutton has a very different architecture in mind and LLMs break a lot of its principles. He calls himself a "classicist" and evokes the original concept of Alan Turing of building a "child machine" - a system capable of learning through experience by dynamically interacting with the world. There's no giant pretraining stage of imitating internet webpages. There's also no supervised finetuning, which he points out is absent in the animal kingdom (it's a subtle point but Sutton is right in the strong sense: animals may of course observe demonstrations, but their actions are not directly forced/"teleoperated" by other animals). Another important note he makes is that even if you just treat pretraining as an initialization of a prior before you finetune with reinforcement learning, Sutton sees the approach as tainted with human bias and fundamentally off course, a bit like when AlphaZero (which has never seen human games of Go) beats AlphaGo (which initializes from them). In Sutton's world view, all there is is an interaction with a world via reinforcement learning, where the reward functions are partially environment specific, but also intrinsically motivated, e.g. "fun", "curiosity", and related to the quality of the prediction in your world model. And the agent is always learning at test time by default, it's not trained once and then deployed thereafter. Overall, Sutton is a lot more interested in what we have common with the animal kingdom instead of what differentiates us. "If we understood a squirrel, we'd be almost done".As for my take...First, I should say that I think Sutton was a great guest for the pod and I like that the AI field maintains entropy of thought and that not everyone is exploiting the next local iteration LLMs. AI has gone through too many discrete transitions of the dominant approach to lose that. And I also think that his criticism of LLMs as not bitter lesson pilled is not inadequate. Frontier LLMs are now highly complex artifacts with a lot of humanness involved at all the stages - the foundation (the pretraining data) is all human text, the finetuning data is human and curated, the reinforcement learning environment mixture is tuned by human engineers. We do not in fact have an actual, single, clean, actually bitter lesson pilled, "turn the crank" algorithm that you could unleash upon the world and see it learn automatically from experience alone.Does such an algorithm even exist? Finding it would of course be a huge AI breakthrough. Two "example proofs" are commonly offered to argue that such a thing is possible. The first example is the success of AlphaZero learning to play Go completely from scratch with no human supervision whatsoever. But the game of Go is clearly such a simple, closed, environment that it's difficult to see the analogous formulation in the messiness of reality. I love Go, but algorithmically and categorically, it is essentially a harder version of tic tac toe. The second example is that of animals, like squirrels. And here, personally, I am also quite hesitant whether it's appropriate because animals arise by a very different computational process and via different constraints than what we have practically available to us in the industry. Animal brains are nowhere near the blank slate they appear to be at birth. First, a lot of what is commonly attributed to "learning" is imo a lot more "maturation". And second, even that which clearly is "learning" and not maturation is a lot more "finetuning" on top of something clearly powerful and preexisting. Example. A baby zebra is born and within a few dozen minutes it can run around the savannah and follow its mother. This is a highly complex sensory-motor task and there is no way in my mind that this is achieved from scratch, tabula rasa. The brains of animals and the billions of parameters within have a powerful initialization encoded in the ATCGs of their DNA, trained via the "outer loop" optimization in the course of evolution. If the baby zebra spasmed its muscles around at random as a reinforcement learning policy would have you do at initialization, it wouldn't get very far at all. Similarly, our AIs now also have neural networks with billions of parameters. These parameters need their own rich, high information density supervision signal. We are not going to re-run evolution. But we do have mountains of internet documents. Yes it is basically supervised learning that is ~absent in the animal kingdom. But it is a way to practically gather enough soft constraints over billions of parameters, to try to get to a point where you're not starting from scratch. TLDR: Pretraining is our crappy evolution. It is one candidate solution to the cold start problem, to be followed later by finetuning on tasks that look more correct, e.g. within the reinforcement learning framework, as state of the art frontier LLM labs now do pervasively.I still think it is worth to be inspired by animals. I think there are multiple powerful ideas that LLM agents are algorithmically missing that can still be adapted from animal intelligence. And I still think the bitter lesson is correct, but I see it more as something platonic to pursue, not necessarily to reach, in our real world and practically speaking. And I say both of these with double digit percent uncertainty and cheer the work of those who disagree, especially those a lot more ambitious bitter lesson wise.So that brings us to where we are. Stated plainly, today's frontier LLM research is not about building animals. It is about summoning ghosts. You can think of ghosts as a fundamentally different kind of point in the space of possible intelligences. They are muddled by humanity. Thoroughly engineered by it. They are these imperfect replicas, a kind of statistical distillation of humanity's documents with some sprinkle on top. They are not platonically bitter lesson pilled, but they are perhaps "practically" bitter lesson pilled, at least compared to a lot of what came before. It seems possibly to me that over time, we can further finetune our ghosts more and more in the direction of animals; That it's not so much a fundamental incompatibility but a matter of initialization in the intelligence space. But it's also quite possible that they diverge even further and end up permanently different, un-animal-like, but still incredibly helpful and properly world-altering. It's possible that ghosts:animals :: planes:birds.Anyway, in summary, overall and actionably, I think this pod is solid "real talk" from Sutton to the frontier LLM researchers, who might be gear shifted a little too much in the exploit mode. Probably we are still not sufficiently bitter lesson pilled and there is a very good chance of more powerful ideas and paradigms, other than exhaustive benchbuilding and benchmaxxing. And animals might be a good source of inspiration. Intrinsic motivation, fun, curiosity, empowerment, multi-agent self-play, culture. Use your imagination.译Sutton（《The Bitter Lesson》作者）在播客中质疑 LLM 并非真正的"苦涩的教训"产物--它们依赖有限的人类数据且充满偏见。他主张 AI 应像动物一样通过 RL 与世界动态交互，而非模仿人类文本。作者认同 LLM 确实充斥人工干预，但认为预训练是应对冷启动的实用"进化替代方案"，纯 RL 在现实世界难以行得通。

Dwarkesh Patel: .@RichardSSutton, father of reinforcement learning, doesn’t think LLMs are bitter-lesson-pilled. My steel man of Richard...

智能体大佬观点数据/训练
推荐理由：Karpathy解读Sutton对LLM的批判，提出「召唤幽灵」vs「构建动物」的深刻比喻，反思AI发展路径

10月1日10月1日周三

星期三 · 1 条
21:36
Sam Altman@sama精选
我明白你们的意思，但是…i get the vibe here, but...we do mostly need the capital for build AI that can do science, and for sure we are focused on AGI with almost all of our research effort.it is also nice to show people cool new tech/products along the way, make them smile, and hopefully make some money given all that compute need.when we launched chatgpt there was a lot of "who needs this and where is AGI".reality is nuanced when it comes to optimal trajectories for a company.译Sam Altman 回应"7万亿治愈癌症却做AI广告"的质疑，解释筹集资金主要用于科研AI与AGI研发，推出个性化广告等消费产品是为赚取算力成本并展示技术，称公司发展路径本就复杂，类似ChatGPT初期也曾遭质疑。

RUNNER TUSHAR: @sama Sam Altman 2 weeks ago: “we need 7 trillion dollars and 10GW to cure cancer” Sam Altman today: “We are launching A...

OpenAI大佬观点
推荐理由：Sam Altman 回应"7 万亿治癌 vs AI 广告"争议，解释 OpenAI 战略平衡逻辑

9月30日9月30日周二

星期二 · 1 条
02:52
Ethan Mollick：One Useful Thing（RSS）精选
真正的 AI Agents 与真正的工作

探讨 AI Agents 在真实工作场景中的定位，指出其核心挑战在于对抗"无限PPT"的形式主义陷阱。强调真正的智能体应当服务于以人为本的实质性工作，而非制造更多文档流程或官僚化产出。

智能体大佬观点
推荐理由：Ethan Mollick 深度解析 AI Agent 在实际工作中的应用与人类价值重塑

9月28日9月28日周日

星期日 · 1 条
13:11
Noam Brown@polynoamial精选
Julian 是 AlphaGo、AlphaZero 和 MuZero 的共同一作，虽在 Twitter 上低调，但十多年来始终处于 AI 指数级进步的最前沿Julian was co-first author on AlphaGo, AlphaZero, and MuZero. He doesn't have a major twitter presence, but he's been at the forefront of AI exponential progress for more than a decade.译AlphaGo、AlphaZero 和 MuZero 共同一作 Julian Schrittwieser 指出，公众对当前 AI 指数级进步普遍认知不足。他发布博客总结近期研究进展，并预测未来 1-2 年的发展趋势。

Julian Schrittwieser: As a researcher at a frontier lab I’m often surprised by how unaware of current AI progress public discussions are. I wr...

DeepMind大佬观点现象/趋势
推荐理由：AlphaGo核心作者谈AI指数级进展认知差与未来趋势，顶级研究员重磅观点

9月23日9月23日周二

星期二 · 1 条
21:45
Sam Altman@sama精选
丰裕智能（Abundant Intelligence）Abundant Intelligence: https://blog.samaltman.com/abundant-intelligence译Sam Altman 提出我们正进入"丰裕智能"时代，深度学习将持续扩展。OpenAI 发布 Deep Research 功能，并预测 2025 年 AI 智能体将加入劳动力市场，未来几年有望实现超级智能，带来科学加速和前所未有的繁荣。
OpenAI大佬观点
推荐理由：Sam Altman 发布新博客阐述 AI 普及化愿景，值得深度阅读
现象与趋势行业动态