← 首页
25SEP
AI前沿每日脉动
AI Frontier Pulse · 中英双语版 Bilingual Edition
2026.09.25 · 周五刊
11 位 Builder23 条推文1 期播客1 篇博客
ChatGPT WorkApplied AIOpen WeightsTokens per WattTrust & Access
Richard Liu · 2026
Curated by Richard Liu · Snapshot follow-builders-2026-09-25-v1
今日头条 · Model And Harness Co-design01 / 15
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.
@trq212X 原文1.7K ❤ · 29 RT · 193 💬
Thariq:lots feedback here, many of you are planning yourself &ampBuilder 动态 · Source Excerpt
Thariq 原文要点:lots feedback here, many of you are planning yourself &amp
Source-grounded excerpt from Thariq: lots feedback here, many of you are planning yourself &amp
@trq212X 原文1.7K ❤ · 29 RT · 193 💬
Guillermo Rauch:Spend of OpenAI vs Anthropic vs Open (Vercel AI Gateway, l低延迟模型 · Source Excerpt
Guillermo Rauch 原文要点:Spend of OpenAI vs Anthropic vs Open (Vercel AI Gateway, last 2 months) • Anthropic still #1 in spend, but went 69% → 40% • OpenAI: 10% → 24% in spend • GPT-6 Astra + GPT 5.6 So…
Source-grounded excerpt from Guillermo Rauch: Spend of OpenAI vs Anthropic vs Open (Vercel AI Gateway, last 2 months) • Anthropic still #1 in spend, but went 69% → 40% • OpenAI: 10% → 24% in spend • GPT-6 Astra + GPT 5.6 So…
@rauchgX 原文635 ❤ · 34 RT · 114 💬
02 / 15
行业分析 · Science, Evals, And Credibility02 / 15
Swyx:more conferences could implement this.Builder 动态 · Source Excerpt
Swyx 原文要点:more conferences could implement this.
Source-grounded excerpt from Swyx: more conferences could implement this.
@swyxX 原文504 ❤ · 29 RT · 45 💬
Thariq:incredibly thoughtful piece on AI and creativity hopefullyBuilder 动态 · Source Excerpt
Thariq 原文要点:incredibly thoughtful piece on AI and creativity hopefully we can make our producst better at enabling and amplifying creatives
Source-grounded excerpt from Thariq: incredibly thoughtful piece on AI and creativity hopefully we can make our producst better at enabling and amplifying creatives
@trq212X 原文518 ❤ · 12 RT · 24 💬
03 / 15
产品设计 · Experience Becomes A Force Multiplier03 / 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.
@steipeteX 原文317 ❤ · 14 RT · 58 💬
Peter Steinberger:Afghanistan is more excited about AI than we are?Builder 动态 · Source Excerpt
Peter Steinberger 原文要点:Afghanistan is more excited about AI than we are?
Source-grounded excerpt from Peter Steinberger: Afghanistan is more excited about AI than we are?
@steipeteX 原文317 ❤ · 14 RT · 58 💬
Peter Steinberger:If you just tell the agent to clean up, it will stop far tBuilder 动态 · Source Excerpt
Peter Steinberger 原文要点:If you just tell the agent to clean up, it will stop far too early.
Source-grounded excerpt from Peter Steinberger: If you just tell the agent to clean up, it will stop far too early.
@steipeteX 原文316 ❤ · 10 RT · 12 💬
04 / 15
设计系统 · Harness, Design.md, Routing, Mobile04 / 15
Josh Woodward:Dreambeans is one of our newer experiments in @GoogleLabs.Builder 动态 · Source Excerpt
Josh Woodward 原文要点:Dreambeans is one of our newer experiments in @GoogleLabs.
Source-grounded excerpt from Josh Woodward: Dreambeans is one of our newer experiments in @GoogleLabs.
@joshwoodwardX 原文149 ❤ · 5 RT · 16 💬
Thariq:this should let you customize the plan mode prompt, createBuilder 动态 · Source Excerpt
Thariq 原文要点:this should let you customize the plan mode prompt, create + share your own modes or just ignore it and rebind shift+tab to something else all together
Source-grounded excerpt from Thariq: this should let you customize the plan mode prompt, create + share your own modes or just ignore it and rebind shift+tab to something else all together
@trq212X 原文137 ❤ · 0 RT · 13 💬
Matt Turck:In Jensen's 5 layer cake analogy for AI (energy, chips, inAgent 安全 · Source Excerpt
Matt Turck 原文要点:In Jensen's 5 layer cake analogy for AI (energy, chips, infrastructure, models, applications), the middle layer (esp the software infra part) is much less understood - and thats…
Source-grounded excerpt from Matt Turck: In Jensen's 5 layer cake analogy for AI (energy, chips, infrastructure, models, applications), the middle layer (esp the software infra part) is much less understood - and thats…
@mattturckX 原文19 ❤ · 8 RT · 9 💬
Swyx:In Jan this year I called my content strategy shot: "ScaliBuilder 动态 · Source Excerpt
Swyx 原文要点:In Jan this year I called my content strategy shot: "Scaling without Slop".
Source-grounded excerpt from Swyx: In Jan this year I called my content strategy shot: "Scaling without Slop".
@swyxX 原文23 ❤ · 0 RT · 12 💬
05 / 15
深度博客 · Official Blog05 / 15
今天的深度信号来自官方博客:AI 工具正在把工作过程转成更可协作的界面。Today's depth signal comes from the official blog: AI tools are turning work into more collaborative interfaces.
Claude for Small Business launches new workflows, integrations, and training programsToday's Deep Blog
今天的博客信号来自官方长文:AI 工具正在把上下文、工具调用和协作流程整合到更完整的工作界面中。
Today's blog signal points to AI tools integrating context, tool use, and collaboration into more complete work interfaces.
06 / 15
播客深度 · Compute Meets The Physical World06 / 15
PODCAST DEEP DIVE
AI 的下一轮竞争不只发生在模型层,也发生在冷却、电网、数据中心、供应链与每瓦 token 数构成的物理系统里。The next AI race is not only at the model layer, but across cooling, grids, data centers, supply chains, and tokens per watt.
Who Feeds the GPUs? Inside AI's Hidden $30B Layer | Renen Hallak, VAST DataPodcast Deep Dive
今天的播客把产品、基础设施和商业化放在同一条线上:AI 不只是模型能力,也是速度、成本和交付系统。
Today's podcast connects product, infrastructure, and commercialization: AI is not only model ability, but speed, cost, and delivery systems.
07 / 15
播客理念 · Industrial Compute At Scale07 / 15
算力上线即被消耗New Compute Is Consumed Immediately
新增容量上线即被工作负载吸收,过早放慢建设通常会带来负面意外。
New capacity is consumed immediately, and slowing the build too early repeatedly creates negative surprises.
关键指标:每瓦 token 数The Metric Is Tokens Per Watt
定制芯片与模型工作负载协同设计,目标是在电力约束下最大化每瓦可生成的 token。
Custom silicon and model workloads are co-designed to maximize tokens generated per watt.
数据中心必须新增电力Data Centers Must Add Power
大规模 AI 基础设施需要同步投资发电、输电、变电站与配套容量。
Large-scale AI infrastructure requires new generation, transmission, substations, and supporting capacity.
算力策略是 All of the AboveThe Strategy Is All Of The Above
多种云、芯片、数据中心与供应路径并行推进,减少单点约束。
Clouds, chips, data centers, and supply paths advance in parallel to reduce single-point constraints.
08 / 15
快讯速览 · Builder Signals08 / 15
Josh Woodward:Dreambeans is one of our newer experiments in @GoogleLabs.Builder 动态 · Source Excerpt
Josh Woodward 原文要点:Dreambeans is one of our newer experiments in @GoogleLabs.
Source-grounded excerpt from Josh Woodward: Dreambeans is one of our newer experiments in @GoogleLabs.
Thariq:this should let you customize the plan mode prompt, createBuilder 动态 · Source Excerpt
Thariq 原文要点:this should let you customize the plan mode prompt, create + share your own modes or just ignore it and rebind shift+tab to something else all together
Source-grounded excerpt from Thariq: this should let you customize the plan mode prompt, create + share your own modes or just ignore it and rebind shift+tab to something else all together
Matt Turck:In Jensen's 5 layer cake analogy for AI (energy, chips, inAgent 安全 · Source Excerpt
Matt Turck 原文要点:In Jensen's 5 layer cake analogy for AI (energy, chips, infrastructure, models, applications), the middle layer (esp the software infra part) is much less understood - and thats…
Source-grounded excerpt from Matt Turck: In Jensen's 5 layer cake analogy for AI (energy, chips, infrastructure, models, applications), the middle layer (esp the software infra part) is much less understood - and thats…
Swyx:In Jan this year I called my content strategy shot: "ScaliBuilder 动态 · Source Excerpt
Swyx 原文要点:In Jan this year I called my content strategy shot: "Scaling without Slop".
Source-grounded excerpt from Swyx: In Jan this year I called my content strategy shot: "Scaling without Slop".
Peter Steinberger:I put Daybreak on it and found 8 more long-standing leaks.Builder 动态 · Source Excerpt
Peter Steinberger 原文要点:I put Daybreak on it and found 8 more long-standing leaks.
Source-grounded excerpt from Peter Steinberger: I put Daybreak on it and found 8 more long-standing leaks.
Peter Yang:This is called stroking the AI's ego and it worksBuilder 动态 · Source Excerpt
Peter Yang 原文要点:This is called stroking the AI's ego and it works
Source-grounded excerpt from Peter Yang: This is called stroking the AI's ego and it works
09 / 15
数据洞察 · Snapshot09 / 15
今日数据概览Today Stats
收录 Builder:11 位
总推文数:23 条
播客节目:1 期
博客文章:1 篇

最高互动:Thariq:lots feedback here, many of you are planning yourself &amp · 1.7K ❤
第二高互动:Guillermo Rauch:Spend of OpenAI vs Anthropic vs Open (Vercel AI Gateway, l · 635 ❤
The snapshot includes 11 builders, 23 tweets, 1 podcast, and 1 blog post.
follow-buildersSnapshot follow-builders-2026-09-25-v1
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.”
@trq212X 原文1.7K ❤ · 29 RT · 193 💬
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.”
@swyxX 原文504 ❤ · 29 RT · 45 💬
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.”
@trq212X 原文137 ❤ · 0 RT · 13 💬
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.”
14 / 15
AI前沿每日脉动
AI Frontier Pulse · 2026.09.25
本期收录 11 位 Builder · 23 条推文 · 1 期播客 · 1 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-2026-09-25-v1