AI 行业热点
🎙️ 播客精选
Google DeepMind’s Logan Kilpatrick: Why the Model Eats the Harness
Training Data · 2026-06-11
Speaker 1 | 00:00 - 00:02
So we could edit this set so it looks like we’re
Speaker 2 | 00:02 - 00:06
here. Okay? Yeah. Yeah. I I want this where where we were talking off camera.
Speaker 2 | 00:06 - 00:36
Like, we should do that for the intro because I think it just, like, makes all this stuff more capable. I’ve seen these examples of, like, such subtle nuance that make me appreciate that it’s like the world understanding playing out. I was giving a talk and was on stage with my friend, Tulsi, who leads the model team. I had mentioned to someone in the crowd to edit the video, and they liter…
🐦 X/Twitter 热点
Swyx (@swyx)
- btw i’ve been shopping around for insurers for the New Media Lab we are setting up (basically the creative playground housing swyx inc) and yeah the NPS of Corgi is insanely high
my real estate broker: “just go with corgi they are covering every single one of my clients rn”
breaking through with ~100% greenfield market share like this is unheard of in the insurance industry [56 ❤️ 1 🔄]
- @QuinnyPig i think this is where i challenge @willccbb to his first poaster session. or maybe @willdepue. or @WilliamBryk. idk all the wills? [7 ❤️]
- @QuinnyPig [6 ❤️]
Thibault Sottiaux (@thsottiaux)
- [2066 ❤️ 67 🔄]
- Now that you can bank usage resets in Codex. Are you a hoarder or do you use them without breaking a sweat? How do you think about them? [2237 ❤️ 27 🔄]
- What should we improve in the Codex app. What’s not delightful? [2581 ❤️ 57 🔄]
Peter Yang (@petergyang)
- It’s funny, growing up as an immigrant I developed a scarcity mindset around saving money and resources.
But when it comes to unlimited token plans and tokenmaxxing, it feels like a missed opportunity if I don’t burn as many tokens as possible and hit the limit each time.
How do I address this disconnect? 🥹 [25 ❤️]
- Check out my friend Kevin’s awesome tool @ferrymanio to post from your favorite social platform and have that cross post to 7+ platforms at once. [12 ❤️]
- Why HTML turned out to be the foundation for agentic video making from @liu8in:
“We’ve been trying to build a video agent. However, we learned the hard way that agents have no visual intelligence.
So that’s when we turned to code. HTML is the LLM’s native language. LLMs can express not only information, but also visual aesthetics through HTML, CSS, and JavaScript.
Footage, images, assets, SVGs can all sit on top of HTML.”
📌 As an example, see how you can make a product video from any website URL here: [46 ❤️ 3 🔄]
Nan Yu (@thenanyu)
- “Quality is irrational” is a great way to describe it. You have to have an irrational level of commitment to constantly choose quality and an irrational level of self-belief that controlling things top to bottom will yield better results than using common frameworks. [69 ❤️ 6 🔄]
- @Outlook @gmail No that’s not going to solve the problem to any significant degree. We are talking about default behavior. [3 ❤️]
Guillermo Rauch (@rauchg)
- The team cooked on performance. ‘Everything the light touches’ was optimized Simba. Painting, layout, WebGPU shaders, blocking scripts. Every frame scrutinized. The best part is that we’ll be updating our with the lessons learned! [791 ❤️ 20 🔄]
- Coding agents will squeeze every ounce of IKEA effect out of you, if you let them. [1335 ❤️ 37 🔄]
- Happy Father’s Day. Thankful to my dad who spent every last penny on buying our family ‘the bicycle of the mind’ [1034 ❤️ 34 🔄]
Aaron Levie (@levie)
- Another new idea to push the state of AI architectures forward. Sakana released a model that effectively uses a mixture of models to get work done. You get a single API but then the work gets farmed out the model that best performs the task.
“Fugu manages model selection, delegation, verification, and synthesis automatically. It solves tasks directly when that is enough, or coordinates a team of expert models when a problem calls for more. The complexity of a multi-agent system never reaches your code.”
This is generally how applied AI products are building their agent harnesses at this point, but the idea of making this an LLM that any developer can interact with is also a great idea. As we get more innovation with both frontier closed and OSS models, there’s going to be a ton of value produced for the layer that can route the best. [206 ❤️ 23 🔄]
- Agents will use software 100X more than people.
When that happens, theres a huge need for guardrails on what the agents are doing so they don’t leak data or change the wrong information, authoritative sources of truth for them to work with, logging and auditing of what they’re doing, the ability to collaborate with people through these systems, and more.
A simple query on any given agentic task could pull in more data than a user touches in a month. As a result, there are lots of categories of software that when it goes headless that the usage and value go up substantially. Agents will end up using our CRM data, documents and corporate knowledge, analytics data, and other information far more than people ever did.
The platforms that can move toward the model of powering these headless interactions, and have a business model and technology strategy to support this, will be in the best position in the future. [228 ❤️ 44 🔄]
Ryo Lu (@ryolu_)
- works with any epub
syncs progress with ryOS account [13 ❤️] - missing wooden shelves
so i made Books in ryOS
started in Cursor mobile, then hand-tuned the animations and textures until things felt right [324 ❤️ 5 🔄]
Garry Tan (@garrytan)
- This is why I made GBrain and open sourced it [88 ❤️ 5 🔄]
- I think one underestimated thing when we look back on it was how useful it is to have your own personal brain and company brain in 2026 at the dawn of usable AGI
AGI gives you the intelligence
You still have to collect your personal context to get the real unlock [964 ❤️ 66 🔄]
Zara Zhang (@zarazhangrui)
- I’m talking about context, not prompt [1 ❤️]
- A good rule of thumb for preventing AI slop (in writing, design, etc):
Is your input (the context) longer than the output?
I’ve found that for the AI to produce quality results, my input is often 3-5 times the length of the output.
If your input is much shorter than the output, it’s almost certainly going to produce slop [131 ❤️ 4 🔄]
Nikunj Kothari (@nikunj)
- How your email finds me today..
Happy Father’s Day to all the families celebrating today 🕺 [57 ❤️]
- Happens every week 🙈
> see an interesting project on X
> play with it, fork it, try it, have some ideas
> go to founder profile
> press the message button
…
> pray to God that I haven’t missed a DM from them before 😢
I try to respond to all the DMs but miss sometimes!! [92 ❤️]
Peter Steinberger (@steipete)
- This is becoming my favorite way to read Twitter. [336 ❤️ 20 🔄]
- People here discussing what happened with OpenClaw.
The hype died down. We improved quality and grew a team. We created a non-profit whereas competiors are VC funded and have other agendas.
This is our strongest week so far. [273 ❤️ 15 🔄]
- I was skeptical about the multi-model routing. Seems my hinch was right. [136 ❤️ 4 🔄]
由 Follow Builders 自动生成 · 2026-06-22