AI前沿每日脉动

23AUG

AI前沿每日脉动

AI Frontier Pulse · 中英双语版 Bilingual Edition

2026.08.23 · 周日刊

9 位 Builder19 条推文1 期播客0 篇博客

ChatGPT WorkApplied AIOpen WeightsTokens per WattTrust & Access

Richard Liu · 2026

Curated by Richard Liu · Snapshot follow-builders-2026-08-23-v2

今天最强信号：GPT-5.6 Sol 只用两项运行设置变化，就在 ARC-AGI-3 达到 SoTA。模型能力不再能脱离 harness、compaction 与跨 context 执行单独评估。The strongest signal: two runtime setting changes put GPT-5.6 Sol at the top of ARC-AGI-3. Model capability can no longer be separated from harness design, compaction, and cross-context execution.

@thsottiaux X 原文1.4K ❤ · 113 RT · 248 💬

Thibault Sottiaux：Update on rate limits in Codex.工作产品化 · Source Excerpt

Thibault Sottiaux 原文要点：Update on rate limits in Codex.

Source-grounded excerpt from Thibault Sottiaux: Update on rate limits in Codex.

@thsottiaux X 原文1.4K ❤ · 113 RT · 248 💬

Guillermo Rauch：San Francisco is a paradise on EarthBuilder 动态 · Source Excerpt

Guillermo Rauch 原文要点：San Francisco is a paradise on Earth

Source-grounded excerpt from Guillermo Rauch: San Francisco is a paradise on Earth

@rauchg X 原文1.1K ❤ · 20 RT · 80 💬

02 / 15

Guillermo Rauch：AI can do many wonderful things and grant many wishes, butBuilder 动态 · Source Excerpt

Guillermo Rauch 原文要点：AI can do many wonderful things and grant many wishes, but we have only one earth, one California, one Patagonia, one Mendoza.

Source-grounded excerpt from Guillermo Rauch: AI can do many wonderful things and grant many wishes, but we have only one earth, one California, one Patagonia, one Mendoza.

@rauchg X 原文735 ❤ · 30 RT · 27 💬

Amjad Masad：“Pretty soon” turned out to be 3 months.Builder 动态 · Source Excerpt

Amjad Masad 原文要点：“Pretty soon” turned out to be 3 months.

Source-grounded excerpt from Amjad Masad: “Pretty soon” turned out to be 3 months.

@amasad X 原文643 ❤ · 43 RT · 37 💬

03 / 15

AI 降低执行成本，却没有消除品味与经验：资深创始人的判断继续复利，设计工具也必须把审美、易用性和乐趣一起产品化。AI lowers execution cost without eliminating taste or experience. Seasoned judgment compounds, while design tools must productize aesthetics, ease, and fun together.

@petergyang X 原文447 ❤ · 23 RT · 20 💬

Peter Yang：If you care about your privacy: 1.工作产品化 · Source Excerpt

Peter Yang 原文要点：If you care about your privacy: 1.

Source-grounded excerpt from Peter Yang: If you care about your privacy: 1.

@petergyang X 原文447 ❤ · 23 RT · 20 💬

Nikunj Kothari：Kids, I don't know what advice you are getting, but ragebaBuilder 动态 · Source Excerpt

Nikunj Kothari 原文要点：Kids, I don't know what advice you are getting, but ragebaiting investors and then sending them SAFE docs to just wire you money is NOT how investing works..

Source-grounded excerpt from Nikunj Kothari: Kids, I don't know what advice you are getting, but ragebaiting investors and then sending them SAFE docs to just wire you money is NOT how investing works..

@nikunj X 原文370 ❤ · 4 RT · 73 💬

04 / 15

Zara Zhang：Everyone who’s ahead in using AI thinks they’re behindBuilder 动态 · Source Excerpt

Zara Zhang 原文要点：Everyone who’s ahead in using AI thinks they’re behind

Source-grounded excerpt from Zara Zhang: Everyone who’s ahead in using AI thinks they’re behind

@zarazhangrui X 原文336 ❤ · 20 RT · 53 💬

Peter Yang：I’m working on a new AI skill.技能化工作流 · Source Excerpt

Peter Yang 原文要点：I’m working on a new AI skill.

Source-grounded excerpt from Peter Yang: I’m working on a new AI skill.

@petergyang X 原文188 ❤ · 5 RT · 39 💬

Aaron Levie：AI diffusion is far more rate limited by having good evalsBuilder 动态 · Source Excerpt

Aaron Levie 原文要点：AI diffusion is far more rate limited by having good evals than most realize.

Source-grounded excerpt from Aaron Levie: AI diffusion is far more rate limited by having good evals than most realize.

@levie X 原文138 ❤ · 12 RT · 25 💬

Zara Zhang：There’s a phenomenon where talented individuals can achievBuilder 动态 · Source Excerpt

Zara Zhang 原文要点：There’s a phenomenon where talented individuals can achieve 10x their potential thanks to AI when working on their own thing But when the same individual is put into a large org…

Source-grounded excerpt from Zara Zhang: There’s a phenomenon where talented individuals can achieve 10x their potential thanks to AI when working on their own thing But when the same individual is put into a large org…

@zarazhangrui X 原文147 ❤ · 7 RT · 14 💬

05 / 15

没有博客并不等于没有深度：今天的深度来自工程、安全、组织与商业化信号在同一天汇合。No blog does not mean no depth: today's depth comes from engineering, safety, organization, and commercialization signals converging.

follow-buildersNo blog post in this snapshot

今日无新增博客：深度信号来自产品与工程现场No Blog Today, But Strong Product Signals

今天中央 feed 没有新增博客，但高互动推文已经形成清晰主线：产品入口、工程基础设施、模型治理和组织能力正在同时变化。

The central feed has no new blog post today, but the top tweets still form a clear arc across product surfaces, engineering infrastructure, model governance, and organization design.

follow-buildersSnapshot follow-builders-2026-08-23-v2

06 / 15

PODCAST DEEP DIVE

模型已经有能力，但产品还没有把能力完整交给用户；记忆、工具与开放协议决定这段 capability overhang 能否被释放。Models have capability that products have not yet delivered. Memory, tools, and open protocols determine whether that capability overhang can be unlocked.

AI & I by Every YouTube 完整播客

Kevin Scott：Agentic Web 需要开放协议Agentic Web Needs Open Protocols

Microsoft CTO Kevin Scott 认为，模型推理能力已经领先于产品交付，形成 capability overhang。要让 agent 真正行动，需要持久记忆、工具访问，以及像 MCP 与 NLWeb 这样简单、可组合、接近 HTTP 与 HTML 角色的开放协议。

Microsoft CTO Kevin Scott argues that model reasoning has moved ahead of product delivery, creating a capability overhang. Useful agents need persistent memory, tool access, and simple composable protocols such as MCP and NLWeb that play roles analogous to HTTP and HTML.

AI & I by Every YouTube 完整播客2026-08-12

07 / 15

产品需要追上模型能力Products Must Close The Capability Overhang

推理能力已经领先于实际产品使用方式，下一阶段重点是把现有能力可靠地交付给用户。

Reasoning capability has moved ahead of product usage. The next phase is reliably delivering ability that models already possess.

Capability YouTube 完整播客

持久记忆支撑复杂委托Persistent Memory Enables Delegation

任务结束后就清空上下文，会让 agent 永远停留在事务型工具；持久、连贯的记忆才能承接更复杂的长期工作。

If context disappears after each task, agents remain transactional. Persistent coherent memory enables more complex long-term delegation.

Memory YouTube 完整播客

开放协议构成 Agentic WebOpen Protocols Form The Agentic Web

MCP 与 NLWeb 的意义类似 HTTP 与 HTML：简单、可组合、足够开放，让 agent 可以连接已有网站、API 与信息系统。

MCP and NLWeb play roles analogous to HTTP and HTML: simple, composable, open layers connecting agents to existing websites, APIs, and information systems.

Protocols YouTube 完整播客

工具不会消灭创造者的手艺Agents Change Craft Without Removing It

软件开发经历过多次工具跃迁；agent 会改变工作方式，但有经验的创造者仍会对过程与结果保持强烈判断。

Software development has changed through many tooling shifts. Agents alter the workflow, but experienced makers still bring strong judgment to process and results.

Craft YouTube 完整播客

08 / 15

Zara Zhang：Everyone who’s ahead in using AI thinks they’re behindBuilder 动态 · Source Excerpt

Zara Zhang 原文要点：Everyone who’s ahead in using AI thinks they’re behind

Source-grounded excerpt from Zara Zhang: Everyone who’s ahead in using AI thinks they’re behind

@zarazhangrui X 原文336 ❤

Peter Yang：I’m working on a new AI skill.技能化工作流 · Source Excerpt

Peter Yang 原文要点：I’m working on a new AI skill.

Source-grounded excerpt from Peter Yang: I’m working on a new AI skill.

@petergyang X 原文188 ❤

Aaron Levie：AI diffusion is far more rate limited by having good evalsBuilder 动态 · Source Excerpt

Aaron Levie 原文要点：AI diffusion is far more rate limited by having good evals than most realize.

Source-grounded excerpt from Aaron Levie: AI diffusion is far more rate limited by having good evals than most realize.

@levie X 原文138 ❤

Zara Zhang：There’s a phenomenon where talented individuals can achievBuilder 动态 · Source Excerpt

@zarazhangrui X 原文147 ❤

Madhu Guru：How to build great evals - part 6 Hill climbing on evals i低延迟模型 · Source Excerpt

Madhu Guru 原文要点：How to build great evals - part 6 Hill climbing on evals is just a fancy way of saying: pick a dimension that matters and optimize for it.

Source-grounded excerpt from Madhu Guru: How to build great evals - part 6 Hill climbing on evals is just a fancy way of saying: pick a dimension that matters and optimize for it.

@realmadhuguru X 原文143 ❤

Amjad Masad：A week has 7 days That means 7 shipsBuilder 动态 · Source Excerpt

Amjad Masad 原文要点：A week has 7 days That means 7 ships

Source-grounded excerpt from Amjad Masad: A week has 7 days That means 7 ships

@amasad X 原文70 ❤

09 / 15

今日数据概览Today Stats

收录 Builder：9 位
总推文数：19 条
播客节目：1 期
博客文章：0 篇

最高互动：Thibault Sottiaux：Update on rate limits in Codex. · 1.4K ❤
第二高互动：Guillermo Rauch：San Francisco is a paradise on Earth · 1.1K ❤

The snapshot includes 9 builders, 19 tweets, 1 podcast, and 0 blog post.

follow-buildersSnapshot follow-builders-2026-08-23-v2

5 条关键洞察5 Key Takeaways

ChatGPT Work 正从任务助手变成可调度上下文、应用与协作的新计算入口。

ChatGPT Work is becoming a computing surface that orchestrates context, apps, and collaboration.

模型越强，企业越需要连接数据、监管、人类决策点与现实反馈回路的 applied AI 层。

Stronger models increase demand for an applied-AI layer spanning data, regulation, human decisions, and real-world feedback.

开放权重从开发者偏好升级为产业领导力与生态扩散战略。

Open weights are moving from developer preference to an industrial strategy for leadership and ecosystem diffusion.

算力竞争的关键约束转向物理世界：电力、冷却、建设速度与每瓦 token 数。

Compute competition is increasingly constrained by power, cooling, build speed, and tokens per watt.

跨应用执行放大了价值，也把 Gmail、日历与办公数据的信任问题推到产品核心。

Cross-app execution expands value while making trust around Gmail, calendars, and office data a core product issue.

10 / 15

本周之声 VOICE 01

只要允许模型持续推理，并借助 canonical compaction 跨多个 context window 工作。

“You just have to allow it to reason and work over multiple context windows.”

@thsottiaux X 原文1.4K ❤ · 113 RT · 248 💬

11 / 15

本周之声 VOICE 02

最好的路径，是赋能科学家，而不是试图由我们自己解决所有问题。

“The best way to do this is for us to empower scientists.”

@rauchg X 原文735 ❤ · 30 RT · 27 💬

12 / 15

本周之声 VOICE 03

我总会先创建 design.md，写清颜色、字体以及其他设计准则。

“I always create a design md file with guidelines for color, typography, and more.”

@petergyang X 原文188 ❤ · 5 RT · 39 💬

13 / 15

播客之声 PODCAST VOICE

AI 是一场“50 年的一夜成功”。

“AI's fifty year overnight success.”

Kevin Kelly · AI & I by Every YouTube 完整播客

14 / 15

AI前沿每日脉动

AI Frontier Pulse · 2026.08.23

本期收录 9 位 Builder · 19 条推文 · 1 期播客 · 0 篇博客

ChatGPT Work · Applied AI · Open Weights · Tokens per Watt · Trust & Access

感谢阅读 · Thank You For Reading

Richard Liu · AI前沿每日脉动 · 2026 · Snapshot follow-builders-2026-08-23-v2