Sam Porter

@Young_AGI K3.1有可以公开的情报吗😋

Try Kimi K3

@claudeai @AndrewCurran_ oh shit

Beginning July 20, Claude Fable 5 will be included in all Max and Team Premium plans, at 50% of limits. Pro and Team Standard users will continue to have access to Fable via usage credits, and will receive a one-time $100 credit. Demand for Fable has been challenging to predict, which is why we rolled it out to subscription plans in stages, extending access several times as we secured additional capacity.

@Young_AGI Cool

Impressive

@Young_AGI Kimi加油啊👍

犯其至难而图其至远者，发之以勇，守之以专，达之以强。🫡 K3

@lidangzzz 进厂罢😭

党哥最绝望的一次，你们讲一下该如何解决

@elonmusk @aaronburnett Is that "Rubin" not "Ruben"?

GPU average power consumption if it’s doing inference over the course of 24 hours is ~2/3 of its peak power, even if you’re super efficient. SpaceX AI Sat V1 peak power spec has been raised to ~250kW (battery-assisted), with average power of ~160kW. Will be able to handle an NVL72 Ruben rack. This is just version 1.

orbital data centers have one glaring constraint few are discussing publicly: inference network coherence. We tackled it this week in our analysis "The Orbital Inference Containment Tax" tbf we mostly hand-waved coherence in our early ODC models (other things to learn first), but went deep after encouragement from someone who's worked on this problem for years. Fair warning: this is the easily the densest analysis we've ever done, because we're modeling at the edge of both AI and spacecraft. If you have the time to dive in, it's really f*ing interesting and insightful for what I expect will happen with orbital architecture over the next few years. If this isn't your day job, share the link with your AI of choice and have it get you up to speed. research.33fg.com/analysis/the-o… core learnings this week: - Frontier LLMs are mixture-of-experts models. Ground operators spread hundreds of experts across big GPU pools down to about 1 expert per GPU. NVIDIA's benchmark shows that spreading is worth up to 1.8x more tokens/second per GPU. and tokens/second is $/second in inference. - A self contained satellite today can't spread that wide. GPUs computing one answer must sync every ~5 microseconds, and light only covers a ~750m round trip in that window. consequently a coherent GPU team ends at the edge. - Your options to address this are formation-fly sats ~150m apart (Google's Suncatcher bet, orders of magnitude closer than Starlinks fly today) or eat the penalty of containing the model inside one sat. - An AI1 sat's power spec ≈ one rack of GB300s (NVL72) 72 GPUs. if you force a 256 expert model into that box then each GPU has to juggle 3.56 experts vs ~1 on the ground. - We conservatively stacked every assumption against orbit and worst case, a sat is 44% less efficient at processing tokens than on the ground. That's what we call 'the containment tax' 1.8 sats worth of GPUs to do the work of 1 note,we used max conservative assumptions as I believe it's important to stress test the orbital compute thesis when reasonable. That said the tax decays, when modeled the levers become pretty clear... 1) SpaceXAI's C-rewrite captures up to half the tax consistent with the token/second uplift we modeled last week. 2) Co-design the model to fit the sat's 72 GPUs. The largest single lever, and it's a training decision, not a hardware program. and 3) add more GPUs per sat via power density. The next node class (assuming a 200kw-240kw class sat) cuts the baseline to 1.44x before software improvements even runs 1 and 3 are already underway at SpaceX. 2 is the cheap, logical next step given their compute advantage and ~3.5-week model release cadence nothing stops them offering co-design as a service to neocloud customers. what we expect? Grok and Composer are the first two models co-designed for orbital inference sats and any external customer co-design comes later. Scoping note: the containment tax hits the revenue and payback side of the equation, not the cost side, so it's not in our orbital-vs-terrestrial cost work yet. These learnings will roll into future iterations of AI compute and the SpaceX Gigamodels. Thanks for reading and happy modeling!

@Young_AGI 杨哥你崛起罢✋😭🤚

✨ An era of bursting inspiration — where ideas ignite and come to life in an instant. The future isn’t coming. It’s already here. 🚀 Are you ready?

@synthwavedd How is it

well hello there

@HaoningTimothy 使用小思考的雷霆大模型可以吗

使用雷霆大思考的小模型并不能降低成本.jpg

@LyingOverDuckww 校友这么强👍

终于要离开这个地方了. 能看到的是荣耀. 看不到的是我的痛苦. 我的精神状态从入校的良好到现在的重度焦虑重度抑郁. 我的困顿我的痛苦我并没有成为我想要成为的人. Algebraist. 我第一天到达上海, 在路上走着, 看到这家咖啡厅, 我的眼角默默地流下眼泪. 我没能成为我想要成为的人, 没有办法.

@elonmusk @aaronburnett C/C++ is great

@aaronburnett Truly massive gains will come in ~3 months when the entire training and inference stack is written in C/C++ and massively simplified (most software layers will be deleted completely) and we exact-map Grok to work incredibly well on a GB300

Let’s see what calling in the aerospace/rocket engineers to work on AI models does for Grok.

@synthwavedd @fanofaliens So it’ll be better than Opus 4.6 and cheaper right

@fanofaliens It was added months ago, they renamed it and launched it as Opus 4.6, and it reappeared today, so no

This post is slightly misleading and old. The exact claude-sonnet-5 slug was reportedly spotted in Vertex AI error logs months ago.

@fiapp_pro @WaitingAlpha 有多会说人话？有Claude的水平吗？

这个只能等 benchmark 了，目前可以确认的是，速度极快（可能因为现在没人用），更会说人话了，前端能力增强，不会莫名其妙打括号说废话，对 threejs 等库的智商程度增强， 代码智能程度大幅增强，例如引用 streamdown 库，会知道自己查看包体积，知道首屏加载的重要性，自动化的进行优化，而不需要你去提示

已用上 GPT 神秘新模型，应该是 5.6 ，大家有什么想测的？

@bridgemindai Source？

Fable 5 access will be restored within the next 48 hours.

@python_xxt 万一这公司真是傻逼呢

人对损失的厌恶感知是得到的2到3倍。 Claude的营销人员，但凡有脑子，都不可能在两周之后把Fable 5真的撤掉。否则，人们会觉得Claude这公司就是个傻逼（国人已经这么认为了），因为那个时候是非理性的，那种损失厌恶，极其上头。

立贴为证， 两周之后，Claude 一定会把fable 5，转为常规供给模型，分配给 Max一档 。

@fkysly 恩情✋😭🤚

A 社：Claude Code 将封禁第三方订阅登录…因为我们的算力紧张，系统资源不足以… Codex：出 Bug 了，重置用量 A 社：发布新模型 Claude Mythos，因为安全问题目前不对外公开使用…Mythos 比 Opus 提升… Codex：周活涨了，重置用量 A社：推出 Agents 管理平台，Notion 都在用，这是我们最新… Codex：昨天时辰不好，明天继续重置用量

@_MegaPolitics Not breaking at all

Breaking: Israeli PM Netanyahu is in Epstein's file

@xiangyuniunai Bro fell into void

@cb_doge powerful and elegant

The evolution of SpaceX Raptor engines.

@corleonescrypto That’s not a real person of course bro

not one dollar moved in 16 years. crazy self control what’s he doing these days?