Agents Over Bubbles | Stratechery by Ben Thompson
Read the Article: https://stratechery.com/2026/agents-over-bubbles/ Links: Stratechery: https://stratechery.com Sign up for Stratechery Plus: https://stratechery.com/stratechery-plus Sharp Tech website: https://sharptech.fm
Read Video · 文字稿与深度分析
本集已有完整文字稿 + AI 深度分析
免费注册 · 无需信用卡 · 注册即获 150 积分,足够解锁本集
- 📄 完整文字稿含时间戳
- ✨ AI 摘要、关键词与思维导图
- 💡 核心要点与精彩引言
- 说话人 1
节目时间轴
Ben Thompson introduces the paradox of AI prognostication and argues that despite bubble fears, the rise of agents means we are not in a bubble.
- Thompson notes a weird paradox in AI prognostication: pressure to take doomsday scenarios seriously on one hand, and pressure to validate bubble fears on the other, though he has long argued bubbles can be good.
- Sitting in March 2026 on the morning of Nvidia's GTC, he has come to a different conclusion: he doesn't think we're in a bubble, which paradoxically may be the truest evidence that we are.
- He frames the episode around three LLM inflection points—ChatGPT, o1, and Opus 4.5—to explain why the industry is compute-constrained and why hyperscaler capex is justified.
The first two LLM paradigms: ChatGPT's flaws and o1's reasoning breakthrough.
- ChatGPT's November 2022 launch opened the world's eyes to LLM capabilities, but its two flaws—frequent errors and hallucinations, plus the need for users to proactively manage mistakes—made LLMs feel like a parlor trick.
- OpenAI's o1 model in September 2024 introduced reasoning over answers before delivery, making LLMs internally proactive about managing mistakes and shifting them from merely useful to reliable and essential.
- Traditional autoregressive LLMs are path dependent: once they commit to a guess they are locked in and doomed to failure, which reasoning models overcome by self-evaluating and considering alternatives.
The third paradigm: agentic workloads with harnesses like Claude Code and Codex.
- Anthropic's Opus 4.5, released November 24, 2025, and OpenAI's GPT-5.2 Codex around December 18 suddenly enabled agents to accomplish tasks taking hours and do them correctly.
- The critical component of agentic workloads is the harness—software that controls the model—which abstracts the user away from the model and lets agents use deterministic tools to verify results.
- In this third paradigm, an agent directs a model to generate code, checks if it works, and retries without user involvement, substantially mitigating the original ChatGPT's flaws for verifiable use cases like coding.
关键概念
- agents— The central thesis of the episode: functional AI agents are the third paradigm shift that justifies massive compute investment.
- AI bubble debate— Thompson argues we are not in a bubble, reversing his earlier position that bubbles can be good.
- three LLM inflection points— ChatGPT, o1 reasoning models, and Opus 4.5 agentic workloads form the framework for the entire analysis.
精选金句
I don't think we're in a bubble, which paradoxically maybe is the truest evidence we are.
💡— A counterintuitive admission that the very act of declaring no bubble could be the strongest sign one exists.
The first flaw is that LM frequently got things wrong and worse would hallucinate when it didn't know the answer. This made LM feel like something of a parlor trick.
🤯— Captures the early perception of LLMs as unreliable novelties, which many bubble proponents still cling to.
可执行的洞察
🤖AI Strategy
Agents require integrated model-plus-harness, not just a good model.
This week, evaluate one AI tool you use and identify whether its value comes from the model or the harness; note how integration affects your workflow.
Enterprise AI adoption is driven by productivity gains, not cost savings alone.
Draft a one-page proposal for your team on how an AI agent could replace a repetitive coordination task, focusing on top-line impact.
📈Career Growth
Agency—the initiative to use AI—is the key differentiator for individuals.
Spend 30 minutes this week learning to use an AI agent tool (e.g., Claude Code, Copilot) for a task you normally do manually.
Fewer people with agency can now drive massive compute demand and economic impact.
Identify one high-leverage project you could execute alone with AI agents and outline a plan to start it this month.
转录文字与 AI 洞察均由模型自动生成,可能存在少量误差。识别效果与音频质量、语速和发音清晰度相关——如有内容看起来不对,以原始音频为准。
播客与视频,已可阅读
音频播客

3 steps to turn everyday get-togethers into transformative gatherings | Priya Parker
TED Talks Daily
2026年8月15日12:38ENPulling Back the Curtain on Sportsbooks & VIP Programs w/ Dillon Borgida | Ep 63
The Risk Takers Podcast
2024年3月14日1:38:10EN
Stock Market EMERGENCY: Sell Your Stocks Now, The Collapse Is Weeks Away!
The Diary Of A CEO with Steven Bartlett
2026年6月25日1:45:21EN
S9E7 鲁豫对话芮迺伟|一生手谈,半生漂泊,归来棋未凉
岩中花述
2026年8月5日1:38:30ZH-Hans
[EP30] Banalidade do mal
QOHÉLET, podcast de Ed René Kivitz
2024年1月18日38:31PT
Martin Luther - Wie wird man Reformator?
Wer wir sind und warum das nicht klappte ...
2026年4月15日1:09:23DE
视频

5 Easy Ways to Make Better Choices Every Day
Daniel Pink
2025年4月28日5:39EN
10 Minutes of This Feels Like 8 Hours of Sleep (Andrew Huberman)
RESPIRE
2024年9月6日9:04EN
QWEN just CRASHED the industry
Wes Roth
2026年8月4日15:21EN
从「上瘾模型」到「专注力训练」,如何在被算法理解的世界里重新找回主动?| 英文访谈 S9E33
声动活泼
2025年10月16日50:48ZH-Hans
Тайная мобилизация уже идет: облавы на улицах и нехватка солдат
Михаил Ходорковский
2026年8月5日7:02RU