Independent Quick Take

We've open-sourced Grok Build and have reset usage limits for all users. Open sourcing Grok Build allows anyone to support making a reliable and robust harness. Check out our code, including the Git repo for the Grok Build CLI. x.ai/open-source

This is an amazing quote from Lindsey Graham about when he realized how politics works. theatlantic.com/ideas/2026/07/…

@CroatiaFooty The sensor did its job. It sensed who the organizers wanted to win and made sure it happened.

HAHAHAHAHAHAHAHAHAHA No touch from a literal cable, but a hair follicle ✅

@ScottPresler We need more of them, faster, with better infrastructure to support them. They are the lifeblood of the 21st century economy; there's a reason why our adversaries want to slow us down. They create American jobs and support/grow American local businesses as well.

What’s your honest opinion on data centers? I want to be able to tell elected officials how you feel about them.

@thsottiaux Continue in cowork*

Excellent stuff. Yesterday I tried doing something in the pop out chat, and I guess it got a bit more ambitious and actionable than I had intended. It actually told me "this would be better in cowork than chat" and to continue in chat. I'm thinking okay cool, I click and it just... Disappeared. It didn't start a cowork task, it just went into the ether, requiring me to rewrite it out. I think the intent was probably to have the user click use cowork and have the task start there rather than just close out the chat window to nothing.

Hello beautiful people! We have reset usage limits across Codex and ChatGPT Work. And another one will come later in the day. Rejoice. Now that I have your attention, a quick update on ChatGPT Work, Codex and all the updates we shared yesterday. We’ve spent the last 24 hours reading feedback, looking at usage patterns, and talking with many of you. The short version is that there is a *lot* of excitement for GPT 5.6 Sol, ChatGPT Work on mobile & web, but also that we didn't get everything quite right. - We made it too easy to use the highest-compute settings without making the impact on usage limits sufficiently clear. - We reorganized the desktop app in one bold move, making familiar things like chats and projects harder to find. - Our launch framing was focused on ChatGPT Work and to some of our Codex fans it made it feel like Codex was going away over time. Absolutely not our intention, we love Codex and it is here to stay. - And we introduced regressions for some existing multi-agent workflows, alongside a collection of rough edges in plugins and other parts of the experience. We’re landing a first set of improvements today. We’re resetting usage twice so people can keep experimenting, changing defaults and the model picker so they don’t push people toward unnecessarily expensive settings, fixing several plugin submission issues, improving how we represent Codex in the product, and cleaning up some of the most immediate desktop problems. A larger set of improvements will land next week. We’re bringing chats and projects back into the sidebar in a more familiar and customizable way, making usage and reset timing much more visible, clarifying when to use ChatGPT Work and when to use Codex, and addressing the many other smaller pieces of great feedback we've had. The ambition behind this launch hasn’t changed. We think bringing ChatGPT and Codex together into a workspace where people and agents can collaborate is a very important step forward. But an ambitious direction doesn’t excuse avoidable confusion or regressions in the first version. Please keep the feedback coming. We’re moving quickly, and you should see the experience already get better with a few updates today; and substantially better again next week.

@aesposito0 @BerntBornich "three questions I'd ask.." *proceeds to not ask a single question*

Three questions I'd ask @BerntBornich as an engineer who worked on the Optimus hand team (and current hand lead at Foundation): 🧵

@kunchenguid "openai's image generation and grok's video generation" Gemini in shambles. But yeah, I dropped Anthropic for OpenAI/xAI and don't regret it.

right now i have fable still in the subscription with limits just reset; i have gpt 5.6 just added to the mix; and i have grok 4.5 also working super well after using all of these for a while, i suspect something really interesting is about to happen tl;dr - i think anthropic is in serious trouble 1. portfolio competitiveness opus and sonnet (who still remembers they have haiku?) are basically not even worth using any more. both gpt 5.6 and grok 4.5 are just outright better fable is the only real competitive option from anthropic now, but it's not even going to stay in the subscription for a while. it's also super slow and expensive not having enough focus on efficiency and being obsessed with intelligence upper bound might have just led anthropic down a bad road 2. compute constraint an interesting side effect of grok becoming so good is that its demand might explode, especially with cursor now heavily doubling down on grok if grok suddenly gets a lot of usage, it will need compute. and xAI may not want to rent compute to anthropic any more if anthropic can't secure more compute, they will be in trouble because there's just no way to grow the business 3. modalities both openai and xAI started investing in multi-modality since the beginning, while anthropic tunnel-visioned into text and code now i find myself increasingly rely on openai's image generation and grok's video generation. if i can only keep one subscription i would have no choice but to drop anthropic because i do need the other modalities overall at this rate, what's going to happen is that the vast majority of agentic work will be done by non-anthropic models, and only very occasionally other models will escalate hard problems to fable as an advisor my advice for consumers remains the same as what I've been saying for a while - get ready for a multi-agent, multi-vendor world. use tools that allow you to freely switch between models, and build a setup where you can use the right model for the right task

@stoolpresidente Yeah but Morocco can't.

Im gonna tell you right now Morracco can beat France

@elonmusk @yacineMTB Not completely wrong (they are a misanthropic bunch)

I was clearly wrong about Anthropic. They are obviously currently the leader in AI. No company has released a model as good as Mythos/Fable and they will undoubtedly have Mythos 2 ready soon. And I would never cut them off in a way that hurt them badly, even as a competitor. That’s not my style. Tesla open sourced its patents and we made the Supercharger network available to all competitors, even though we could have made it a walled garden. SpaceX launches competing satellite systems with no increase in price or use of unfair terms. Even my worst enemies can attack me on this platform. …

SpaceXAI now has a legitimate frontier model that competes with opus 4.8. Also, Anthropic is completely reliant on the compute rented from SpaceXAI If Elon wanted to kill anthropic, he could. Iirc the compute lease was short term, 6 months from May without renewal promise. GG

I could tell the difference within two prompts. Well done over there. Looking forward to using this. Also, the ChatGPT app that replaced Codex, if I could make a suggestion: projects from the chat side should be migrated to the app. Right now any old project folders I had from brainstorming on ChatGPT (not on Codex) is not in the app. I can bring up the chat thread itself in history for chat, but that's very disorganized since it's by recency. As somebody who puts a premium on staying organized, it would be helpful to keep those chats organized while using the new app. (If it isn't in a project folder, it'll get lost and I can't refer back to it easily. And you can't search chats in the desktop app either. So double the disorganization.)

GPT 5.6 Sol is our best model so far at ~everything. It's remarkable how stark the difference is with GPT-5.5 across coding, hard context and strategy tasks, building sites, ... and literally anything I need to get done. Totally requires raising the ambition of what you try.

I don't know why more people don't see it. I don't know if people recognize that X is owned by xAI, which is owned by SpaceX, who also owns Cursor. And starlink, for high speed internet anywhere on earth. All of which is run by the same person that runs driverless vehicles, cutting edge battery production, and soon to be humanoid robots. Oh and they are jointly going to build terafab, to make current processor manufacturing look like a dripping faucet with the output they're aiming for. You aren't buying SPCX for the rockets is what I'm getting at.

One of the more interesting Grok bull cases I heard at ICML was this: The core idea is that xAI may actually be better positioned than OpenAI Codex in the coding-agent market. The reason Claude Code is currently leading in coding agents is not just model quality. Claude Code effectively pioneered the category at scale, which gave it one of the largest user pools in the industry. More users mean more real-world coding data. That data can then be used to improve Claude Code’s quality, which attracts even more users, creating a flywheel of more users, more data, and a better product. Seen through this lens, xAI’s acquisition of Cursor starts to make a lot more sense. Cursor likely has a much larger real-world user base and coding dataset than Codex. If xAI can effectively train on and leverage that data, the argument is that overtaking Codex may only be a matter of time. I just saw some reactions to Grok 4.5 on my X timeline, and the feedback seems more positive than I expected. Is this thesis starting to play out?

@Dispropoganda Oh, so you're that level of stupid.

@DebLaughton You guys showed up after the Soviets did all the hard work and took all the credit.

Dear Americans, this is how much of Europe it took to beat you in the world cup in your own home. Try being a little bit more humble next time.

@usmntonly Oh well. At the end of the day, we still won, for the simple fact we don't have to live in Europe.

Belgium players did the Trump dance in the dressing room to mock the USMNT after their 4-1 win in the World Cup Round of 16.

@stoolpresidente He shouldn't have went to ground there. It wasn't necessary and created a dangerous situation when there didn't need to be one. Yes in this instance, studs to the calf, hurts. But I've seen that sort of tackle turn into a season ending injury if the foot is firmly planted.

I hate that call. His leg slid off the ball. A red should be malicious.

@erling Great stuff today! Glad I got to see it in person.

Worth the 28-year wait! 🇳🇴❤️

@bereketB44 @rodger You mean that group stage game where winning literally made no difference to winning the group so we benched a bunch of starters? That loss to Turkey?

@rodger Ok, so USA lost to Turkey and Paraguay beat Turkey... you see how that does not work?

Again, if I were France, I would simply score four goals (three in the first half) against Paraguay. I was told this was easy when the United States men’s national team did it

@peterrhague People these days just call it "journalism".

Do we have a name for this kind of obfuscation yet?

FABLE 5 CAME BACK NERFED. We re-ran the July 1st version of Claude Fable 5 on BridgeBench. The results are brutal: Debugging: 86.2 → 25.9 Refactoring: 73.6 → 38.4 Hallucination: 75.9 → 61.7 The new guardrails are kicking in on way too many tasks and falling back to Opus 4.8. This is not the model that got banned. Anthropic owes everyone an explanation.