The thing I'm learning about business nowadays: to survive, you need to constantly reorient. You used to build a business, hit product-market fit, and then kind of coast. Now you hit PMF and the ground shifts 3 months later. New AI models, new algorithms, new tools that change what's possible. So you reorient, and reorient again, it's pretty constant. I'm noticing even companies doing $100M to $1B in revenue are sweating these days. They feel it too. But the cool part is if you reorient right, you get rewarded massively. That's kinda the trade. The stability is gone, but so is the ceiling. I've been reorienting my own companies more this year than in the last five combined. And weirdly, it's the most alive I've felt building. It's a good time to be a founder if you can move on a dime. For the experimenters out there, the people who actually like the constant change... I think this has gotta be the greatest time to be building there's ever been.

babe. futures are up. i feel it. the candle this week is going to be green and it’s going to be huge. you're gonna be like: "please, please, we can’t win anymore. please, elon, we beg you sir, we don’t want to win anymore. it’s too much chili's. it’s not fair to everybody else."

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.

I've reached 900k followers, thank you everyone! Just some reflection, it's actually a nightmare at higher counts... Which is new to me, since I started at almost nothing earlier this year. 1. Threats of safety, which I don't enjoy. - Random new accounts sending personal threats - Mysterious foreign accounts, sending coercive messages the moment I mention a foreign ticker If I hold an opinion about a stock, I'd prefer not to have anyone show up on the street with a knife at night since it's safety risk at this point... Never felt like I'd have to take that into consideration for holding an opinion. 2. Egregious media narratives - Lot of outlets try to dismiss my ideas like $RPI or $SIVE as "meme stocks" or me as "meme trader" when I talk about supply chain chokepoints or fundamental catalysts to revenue due to AI. Without ever covering the core ideas or hyperscaler mapping I've done. I let success validate my ideas, but it's annoying when they can just pretend their claims like "___ will imminently crash as a fact" which ends up not happening. Or that "___" is just a retail meme stock without mentioning $JBL, Ayar, $GFS or others for CPO scale up or pluggables. I'm grateful to some outlets, like in Belgium for $XFAB coverage, but many others try and paint their own narrative rather than covering a very nuanced thesis objectively. - When I mention a ticker in a foreign market, some try to frame me as some "adversarial account"! I'm just sharing ideas, and I'll always disclose positions/financial interest if I have them. If I say I have none, they just make up some narrative that I'm some foreign institution or random group inside their country that does. Some have even tried hilariously forging screenshots of me to harm my reputation, which was annoying to deal with. Since it's easier to textedit a photo and spread it around than it is to dispel disinformation. Lot of institutions know me personally by now. I just don't go broadcasting my identity to random people in the media/public, since personal security gets expensive and I just want to live about my day. 3. Endless X bot spam - I have endless hundreds of comments a day trying to link external communities pretending they're me. I manually delete, and block thousands of these accounts, sometimes hit regular people by accident. And it's extremely exhausting to the point my fingers cramp certain days. Even if I give the most blatant warnings in bio, people still get scammed and think it's me. Then I get blamed for it too. I actually had a conversation where someone asked if a scammer was me. I said no verbatim, then they double checked like 2-3 more times because they didn't believe I said no because someone had the same username on another scam platform. In another world, maybe we'll get automod type tools. But until then, it's mentally exhausting dealing with all of this. But I do so anyways since I don't want my followers to get scammed. 4. Endless engagement baiting - I have tons of people making up fake stuff about me, tagging me in posts, randomly bashing my ideas while missing any technical nuance in order to get me to comment or farm views. It's a catch 22 since the moment I reply, they get a lot of views, which was their goal. And if I don't reply it just looks like I'm avoiding something (which is not true). I'm very transparent, I always try and entertain people who disagree first. But when it gets into personal attacks or leading questions/disinformation rather than substance then I don't engage. It's a little sad though when people I've interacted with end up caring more about engagement over relationships down the road. 5. People missing the nuance - I share ideas pretty often and I always maintain I don't want people to copytrade or blindly follow, which is why I haven't set up any copytrading apps. Before it was just posting with a few $RDDT friends on ideas like $AXTI, $NBIS, or $ALAB, and people were more knowledgeable traders. Then seeing if markets validate my guesses like with InP substrate bottlenecks. But because many turned out directionally right, to the point it became a central talking point between China/US trade relations, lot of random folks started to follow my account last few months. But many of them miss nuances with float dilution, or material updates along the way such as rate hikes, and I get blamed if a thesis changes. Now since there's so many random people, even if I share an idea about $CBRS last Friday. At market open I had people panicking over a 2% drop, asking me if I did something to a $44B+ company, even if it went up 20%+ today. Now I have compliance review for posts, which delays things a lot but makes sure people don't get the wrong idea. And I have to be careful about my timing, so I post new ideas when markets close. Before, it was just sharing ideas the moment I had them during lunch or in the shower, so this is becoming a headache. 6. I post my core ideas for free. And I've always said this from the start. I actually made my subscribers chat literally by accident, and set the bare minimum amount on purpose since I didn't want to take people's money. But kept it so I could avoid the bots and not spam the main timeline with random thoughts/scrapped research. I think me becoming the #1# most subscribed to person on X was literally by accident. But lot of weird narratives out there when media or others don't believe that I just genuinely want to share my ideas for free. And that people shouldn't need to pay $300 in a paywalled community to see them. _ Not sure if people want to read it all, but just some things I've gone through. It's not as much fun at higher numbers and it would be nice to go back to when I was smaller. Regardless, I'd rather not turn this into a job, or make it feel like one. Will probably take a break when I hit 1M as a milestone and do something else that will help out the community like sharing my research pipelines and such. Think a lot of my friends at 10-50K follower counts, seem to experience some of these issues too. But problems get magnified a lot at this level. I never needed the X creator fees or anything, just wanted to share ideas for fun if people wanted to listen. And didn't expect any popularity, but it's nice to see that people find my ideas compelling.

Around 1900, American factories started replacing their steam engines with electric motors. Funny enough, economists later found it took ~30 years for electricity to show up in productivity numbers. Not because the technology didn't work. Because the building was still shaped like its old power source. The gains arrived when a new generation rebuilt the factory floor around small motors: one per machine, arranged around the flow of work instead of the shaft. @ArmanHezarkhani and I keep circling a question that gets at a similar idea: what do products look like when the primary user is an agent and the human comes second? It is hard because every form factor we own, the computer, the phone, the app, was built human-first. - Screens exist to translate machine state into something eyes can parse. - Buttons exist to translate intent into something machines can read. - An interface is a translation layer between human senses and machine memory. Agents do not need the translation. They read the machine state directly. So when we watch an agent drive a cursor around a webpage today, we are watching the factory with the new motor bolted where the steam engine used to be. It works. It is also a sign the building is shaped wrong for its new occupant. The clearest way I have found to think about what comes next: product design stops being one surface and splits into three. 1) The intent surface. Human to agent. How you express what you want. This surface is shrinking toward speech and glances, because stating intent takes a sentence, not a screen. The demand for big glass came from humans doing the manipulation themselves. Move the manipulation to agents and the hardware can collapse into ears and pockets. 2) The execution surface. Agent to world. Protocols, APIs, MCPs, tool registries. Humans never see this layer, but it is where most of the product now lives. Today the API is a side door and the interface is the product. That inverts. The protocol becomes the front door, and the pretty interface becomes one optional rendering of it. 3) The verification surface. Agent to human. How you check the work and feel safe approving it. I think this is the most underrated of the three. Trust is the bottleneck of the whole system. The products people love will be the ones that make "show me what my agent did" feel effortless: previews, diffs, undo. Approval becomes the core gesture of computing, the way the click was for thirty years.

You can feel the problem yourself Text someone about a product, a place, or something you want to buy on WhatsApp. Then open Instagram or Facebook… and suddenly the ads feel way too relevant Maybe it’s metadata. Maybe it’s activity tracking. Maybe it’s cross-app profiling. Or it’s all of the above The truth is, you can’t fully trust Meta with your privacy. There are many active lawsuits and years of scandals surrounding their claims Using WhatsApp is like giving the keys to a robber to your personal vault Try XChat It is built from the ground up with privacy first. No ads. No tracking hooks or ad profiling and it's fully End-to-end encrypted It’s the messaging app you can actually trust Ask your family to join it too

Someone built a free tool that makes Claude Code burn 10X LESS TOKENS AI coding agents start every session from zero without memory of your codebase The agent searches file after file and burns thousands of tokens just figuring out where things actually are, then next session it repeats all of it again You pay for all of these tokens and none of that is actual work Codebase Memory MCP takes a different approach It reads your codebase once and builds a permanent map of it, every function, every connection, who calls what and what depends on what From that point the agent pulls answers from the map instead of reading files to find them Questions like what calls this function or what breaks if I change this file get answered from the map instead of a 20 file read Same 5 questions that cost you 412,000 tokens without using it now ONLY cost 3,400 using the map This tool mapped the entire Linux kernel, 28 million lines of code, in 3 minutes and answers questions about it in under 1 millisecond The problem scales with repo size, small projects barely notice anything, while big codebases feel it immediately and on flagship models it gets even worse Claude Fable charges $10 for every million tokens it reads and GPT 5.5 half that, on Fable those 5 questions cost over $4 or roughly 3 cents using the MCP This is a big reason people downgrade to cheaper models mid project Memory tools for coding agents already exist but three things separate this one: - Most tools in this category run a second AI model to build and search the memory, another key, another bill, while this builds the map with pure parsing and costs zero tokens to run - Tools that use AI summaries of your code are outdated the moment you edit a file, this one rebuilds only what changed in seconds so the map always matches your code - Every MCP shares the same weakness where the agent forgets it exists and starts searching in the same ineffective way again, this one catches those searches and slips the map into the results so it works even when the AI ignores it The map itself is a single file every team can easily commit to the repo and everyone who clones it starts with the map already built Everything arrives as one small binary that knows 158 programming languages and there's nothing else to install It works with 11 coding agents including Claude Code, Codex and Gemini, all of it running locally so your code never leaves your machine The creator even tested it on 31 real codebases for an arXiv paper and the agent finished the same work in half the steps This project is only five months old with 30,000 stars and 35 updates and right now it sits at the top of GitHub's monthly trending

Tesla road trips are starting to feel different. FSD handles the driving. Grok handles the curiosity: "Hey Grok, what's that mountain?" Ask about anything. Geology. History. Landmarks. Local stories. Nearby attractions. You drive through a place once. Now you actually learn something about it too. Having an xAI tour guide permanently riding shotgun is a surprisingly cool future. @Tesla @Grok