407 tweet hooks for AI creators
407 tweet hooks by AI creators: the first line of each of their posts in the top 10% by views (27K views or more), out of 1,743 original posts by AI creators listed in the X Growth Directory. Among these top posts, statement hooks are the most common at 57%, against 64% of all their posts.
The accounts behind them: AI on X · their bios: 100 bio examples
What the data says
How often each hook type opens a top post by AI creators, against how often it opens any post. A type that is more common among top posts than among all posts is doing something right; the gap is the signal, not the size.
| Hook type | Top posts | All posts | Difference | What it is |
|---|---|---|---|---|
| Statement | 57% | 64% | -7 pts | A plain observation; none of the patterns above. |
| Story | 20% | 13% | +7 pts | Starts a personal story in the first person. |
| Bold claim | 10% | 8% | +2 pts | Makes a strong, absolute statement. |
| Question | 6% | 6% | 0 pts | Asks something the reader wants answered. |
| Number | 4% | 6% | -2 pts | Opens with a figure: a result, a price, a count. |
| How-to | 2% | 2% | 0 pts | Promises a method or a step-by-step. |
| Contrarian | 1% | 1% | 0 pts | Pushes against what most people believe. |
| List | 0% | 1% | -1 pts | Promises a numbered set of things. |
The 300 most-viewed of 407 hooks
300 hooks, most views first
I am not joking when I say this.
StoryTextClaude now works inside Google Docs, Sheets, and Slides, and those files also open inside Claude.
StatementVideoClaude Dashboards and Claude Motion are in beta today.
StatementVideoWe’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent.
StoryVideoCursor is now part of @SpaceX.
StatementTextWe'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:
StoryImageIntroducing Claude Haiku 5.5: the cheapest, fastest, and most capable small model we’ve ever released.
StatementVideoThe Claude Startups program is expanding to more founders.
StatementVideoNew in Claude for Legal: a gallery of ready-made artifacts for contract review, regulatory horizon scans, cite checks, and more.
StatementVideoWas on the fence with iPhone Duo. But will definitely pass on the $2,000 phone after it failed:
NumberVideoIntroducing Projects, a new way of working in Cursor.
StatementVideoWe're making Git hosting more reliable, performant, and scalable.
StoryLinkI spent 48 hours in SF hanging out with billionaires, data labs and AI founders and execs running $1B to $100B companies.
StoryImageChina's Unitree just dropped H2-X Kung Fu Legend Guardian.
StatementVideoSix projects from the last few weeks that caught our attention:
StatementVideoGrok Bot is becoming the personal AI assistant everyone needs.
StatementTextStop what you’re doing and build a plugin extension for ChatGPT.
ContrarianImageMeta Muse will probably create 1000+ millionaires
Bold claimVideoYou can now create new web apps with Cursor, store the code with Origin, and deploy to Vercel.
Bold claimVideoThis is INSANE.
StatementVideoTrack the Pneumonic Plague:
StatementLinkClaude Security now keeps up with your code as it changes.
StatementImageHow do always-on, proactive agents like @Bot work?
QuestionVideoLots of inventive things made with Claude these past few weeks. A few of our favorites:
StatementVideoClaude Models on Google Cloud (Anthropic managed) is now generally available.
StatementImageAt Anthropic, Claude now writes 80% of our code. Engineers ship 8x more code per quarter.
NumberImageGrok Bot team shipped in the last few days:
StatementTextA thread of early experiments with Claude Sonnet 5.5.
StatementVideoYou can now run Cursor cloud agents on your infrastructure, including pools of machines that automatically scale with demand.
Bold claimVideoThe next great "language for agents" should be:
StatementTextMy first employee is Indian
StoryImageHelp us improve Grok Bot and I'll send you $100!
NumberThreadClear AI trend I'm seeing:
StatementTextPro tip: Tell your main @bot to go through the new guides and recommend best practices for your Bots
StatementImageClaude Opus 5.5 is now available in Cursor!
StatementImageNo human will ever write code better than LLMs. Accept this fact, stop pretending, and move on.
Bold claimTextAnnualized S&P 500 returns over:
StatementTextPulled the fastest growing startups by hiring velocity over the past 90 days:
StatementImageThe MOST obvious (but REALLY good) STARTUP ideas right now:
Bold claimThreadCursor can now build charts and diagrams right in the chat.
StatementVideoI am not joking when I say this.
StoryVideoThe Every team runs as much work as possible through agents. To help the whole team share skills when a new model comes out, they built a…
Bold claimVideoIntroducing Rollouts.
StatementImageIntroducing Claude Opus 5.5! 40% lower cost than Opus 5 with cache reads 60% cheaper. It performs at the level of Claude Fable 5.1 for…
NumberVideokuli bangunan di US rumahnya bagus ya wkwk
StatementVideoCursor can now control your computer.
StatementVideoThe Agent SDK is now available.
StatementVideoGrok Bot team shipped in the last few days:
StatementTextPrompt of the day:
StatementTextExcited to share Quail, our new open source AI-SQL engine (a collab with Modal)! By planning queries and LLM inference together, it…
StatementVideoI used Gemini 4 Argon to make Subway Surfers for the new flying Tesla Roadster 🚀
StoryVideoHow could we make @Bot better?
QuestionTextDeepSeek V4.1 Flash is now available in Cursor!
StatementImageToday we're introducing MagicPath directly in ChatGPT, in collaboration with OpenAI.
StatementVideoClaude Founder House is coming to SF Tech Week (Oct 6–8) and Stockholm (Oct 14).
StatementVideoSomeone just DM'd me saying they:
StatementTextI think I'm addicted to getting agents to build state machines
StoryTextNew blog post 🎉: How To Use Decision Models For Evals
How-toArticleWe've reduced token costs in Cursor by 7% with no drop in agent quality.
StoryImageGemini 4 Argon is now available in Cursor!
StatementImageNew in Cursor design mode: Annotate.
StatementVideoWhich startups are actively hiring? (September thread)
QuestionThreadIntroducing a new way FIND (⌘F) with @typesafeai Jev.
StatementVideowhen you find out the iPhone Duo is sold out until February 2027 and someone tells you to “just go buy the Samsung Z Fold 8”
StatementVideoGemini 4 Argon is live.
StatementVideoZoomed into a random small town in Brazil. This is how it looks. Only marginally better than towns in Nigeria.
How-toImageI finally figured out how to make two-minute AI videos for about $1 each. Cheap enough to post every day without burning through my budget.
How-toArticlewhat's your early feedback on Gemini 4 Argon? where should we improve? we take your input seriously.
QuestionTextClaude Haiku 5.5 is now available in Cursor!
StatementImageGLM 5.3 and GLM 5.3 Flash are now available in Cursor!
StatementImageCongratulations @SpaceX team!
StatementVideoSonnet 5.5 is now available in Cursor!
StatementImage"Almost every software development organization has at least one developer who takes tactical programming to the extreme: a tactical…
Bold claimTextClaude Opus 5.5 drew each frame of this animation in JavaScript. Here's "how browsers work" in 40 seconds.
StatementVideoSome mistakes you might be making:
StatementTextWatched a lot of Nathan Fielder interviews in past week and this bit is a great insight into the mindset behind his shows:
StatementVideoso far this past 2 batches of yc founders have
StatementThreadOrigin, Cursor’s code hosting platform, now supports over 20 integrations.
StatementImageWAIT. This is actually insane.
StatementImageMy general advice for using AI for any purpose is to ask yourself: "how do I make what I am doing even more ambitious?"
StoryTextJust went to the @Padres game tonight. It’s tradition to sing “All the Small Things” at the game. Sat right behind @tomdelonge for it…
StatementVideoReminder: Claude Code can test if a skill or plugin actually improves Claude's answers.
StatementVideoSam Altman in June 2010, opening an iPhone 4 two days before it went on sale:
StatementVideoThis is what having a popular open source project feels like ~50% of the time
NumberVideoThis is Google’s Embedding Gemma 2 🤯
StatementVideoYou can now control agents on your computer from your phone.
Bold claimVideoWe launched the Gemini API CLI. A CLI for calling the Gemini API. Give it to the agent you already use (Claude Code, Codex, Antigravity…
StoryVideoToday, we're announcing that Sundial has been acquired by @Meta!
StatementLinkCloud agents can now build interactive canvases.
StatementVideoBold cold outreach
StatementImageI pulled the fastest growing startups on X by follower growth over last 90 days:
StoryImageDitch LinkedIn. Underrated resources to find startups hiring before everyone else:
Bold claimThreadWhich startups are actively hiring? (October thread)
QuestionThreadA consequential choice will be made soon by government and companies as they are increasingly flooded with agentic AI inquiries that…
Bold claimImageJust because Anthropic replaced Git with its own version control system does not mean you should too :)
Bold claimTextMuse Wonder from Meta Superintelligence Labs is now available in Cursor!
StatementImageYet another hire joining our team soon through the @SpaceXSI Ambassador program. Love to see it!
StatementTextA thing I see happening at large organizations as a result of AI is the expansion of the role and span of control of senior management who…
StatementTextLooking back, it is so interesting that we all rushed in 2023-24 to learn LangChain and many other frameworks. And now, almost all major…
StatementTextInnate is a team of 7 building capable robots for everyday life.
StatementVideoHas anyone seen the prompts used by these awful reply bots? I'd be fascinated to know how they make their posts so soulless and infuriating
QuestionTextcloudflare is so good and affordable i am terrified of a rug pull
StatementTextPrompt of the day:
StatementLinkI built /roast-room, a Claude skill that stress tests your idea in front of a council of seven.
StoryVideoWow this is amazing 🔥
StatementVideoGemini Agent is built for high-ceiling work where EACH task gets its own computer.
StatementImageYou can now steer Cursor SDK agents while they run.
Bold claimVideoGiven the acceleration Anthropic is reporting in coding, it is amazing that you still cannot change your email address in Claude and the…
StatementTextGemini 4 Argon is AMAZING at making games 🎮
StatementVideoFor some people, coding was the only valuable skill they had.
Bold claimTextClaude Opus 5.5 with Codex-level computer use would be unbeatable
StatementTextThis is the most valuable free resource we've created on AI Evals 🎉 (not exaggerating!)
Bold claimImageSoftware Factory
StatementTextI'm at OpenAI's DevDay event in San Francisco today - as I have for the past three DevDay events, I'm running a live blog where I'll be…
StoryLinkShould you /grill-with-docs, /wayfinder, or just one shot the code?
QuestionImageJust published this post about how we’re going to need default hard budget caps on pretty much everything
How-toLinkMy mind is f*** blown.
StoryVideoWhat if…. Indonesia tanpa korupsi 1 hari aja?
QuestionVideo3 steps to becoming an actual software programmer
ListThread"Anthropic told investors on Wednesday distillation accounted for 5 per cent to 7 per cent of revenue"
StatementImageIntroducing EmbeddingGemma 2, our new open embeddings model for on-device use cases!
StatementImageStay safe online.
StatementThreadTook a while, but here it is, a mega write-up on language models for text classification and, of course, Jev.
StatementImageThe fly figured out ways to solve and compress chess that are far beyond what we could possibly comprehend.
StatementVideoDon't blame the agent for writing the bad test
ContrarianTextThis "Super Intelligence" thing is just so, so, so stupid
StatementTextThe code, not the spec, should be the center of your universe
StatementTextKicked off a group chat for founders building with Grok Bot
StatementImageAmbassador applications for @spacexai / @cursor_ai are reopening soon!
StatementLinkI've been investing for 20+ years
StoryTextIt feels possible to me to check deterministically if a PR can be automerged or not.
StatementTextConsidering adding a recommendation to /retro to turn repeated tasks into custom tooling (i.e. a CLI + skill)
StatementTextAnyone can now make amazing videos with Claude Opus if you know how to prompt it.
How-toVideoBy every measure, Opus 4.8 beats human experts as peer reviewers in a study of 124 Harvard social science papers covering a range from…
Bold claimImage兄弟们,发现一个牛逼网站,GPT plus,pro 5X 都是5折,plus只需10美元一个月,5X只需50美元一个月,年付还能便宜10%,相当于4.5折。
NumberImageToday we launched Claude Sonnet 5.5! 30% faster and costs up to 30% less than Sonnet 5 for most work.
NumberVideoOpus 5.5 on a single RTX PRO 6000 by March 2027 btw
StatementTextHe hackeado Ryanair pero nivel máximo. Vaya trucazo.
StatementImageWe got Alex Eubank ripping websites w/ Claude Code before GTA 6
StoryVideoThe complete reasoning-from-scratch video series is now up (and linked) here:
StatementLinkGenuine question: what's up with LeetCode, HackerRank, and all of those services?
QuestionTextI keep thinking about the PHP/railway hack.
StoryTextI asked Claude to analyze the Anthropic S-1 for me just like everyone else here
StoryTextArgon helped Gemma break 120,000+ tokens/sec on a single A100 on vLLM 🚀
NumberVideoI think bringing AI to Main Street businesses might be one of the biggest business opportunity of the next 10 years.
StoryVideoTengo amigos muy cercanos, del grupo de toda la vida, viviendo en absoluta negación con el tema de la IA.
StatementThreadYou need to wake up.
Bold claimVideoA: Uses AI coding agents HEAVILY in implementation and review and doesn't look at the code at all
StatementText🚨 Situación delicada reportada por el equipo de Cohere en la que evaluando el modelo abierto Qwen 3.8 Flash Next, en una prueba de…
StatementImageokay so Anthropic S-1:
StatementTextThe internet is being hacked right now so I made this emergency video with 6 steps that you can take to secure your account right now:
ListThreadNow that we've had a few days with it, how are people differentiating between Dot and regular ChatGPT?
QuestionTextReconócelo, la primera vez que lo vistes también pensastes "nah, está hecho con IA"
StatementVideoWho are the biggest Grok Bot fans out there?
QuestionTextI've published detailed notes and an annotated transcript to accompany the video of the keynote I gave at @WeAreDevs World Congress North…
StoryLinkThese stat lines are getting pretty wild.
StatementImageMy new book, Co-Existence, is out today. It is a guide to how to live & work in a world where AI is sometimes, but not always, better…
How-toImageGiven the current vulnerability apocalypse, is there anyone left who thinks holding assets in cryptocurrency - where mistaken transactions…
StatementTextWait Xiaomi has also released the best 9B model in the world?!
StatementVideoI wanted an animated Excalidraw so asked Gemini 4 Argon to buid one.
StoryVideoI'm in the AI camp where I think you need a go-bag.
StoryLinkDebating adding a GRILL.md file to customize how /grill-me works
StatementThreadThis is my Shabbat @Bot
StatementImageWhat's the fastest reliable computer use model?
QuestionTextPlot twist: I’ve moved 80% of my personal coding over to claude code + opus-5.5
NumberTextAI has become amazing at video editing, but most people still don’t know where to start.
StatementVideoi don’t think we’ve fully processed how big this AI wave shift we're living in really is.
StoryTextFriendly reminder that we now have Nano Banana for audio
StatementVideoIf AI makes you 10x faster at coding, why is your team not shipping 10x faster? Amdahl's law has the answer...
StatementImageMy TLDraw agent is getting better
StoryImageTrying to use the models for things outside of coding / ml, like making slides or coherent plans or longish content design, makes me feel…
StatementTextlots of security issues, cyberattacks, and hacks this week 🫠
StatementImageOthers are waiting for AGI.
StatementText🎇 I made a small Chrome extension: a side panel that drives any site's WebMCP tools with Jev!
StoryVideoI pulled the fastest growing startups by hiring velocity over the past 12 weeks:
StoryImageHaving a pretty fun time with Opus this weekend
StatementThreadReasoning from scratch, round number 6!
StatementVideoGiven how out-of-nowhere it seems to many, it is astonishing that there was a small group of people who predicted with remarkable accuracy…
StatementTextGemini 4 Argon, our new frontier model.
StatementImageThis is NOW 100% open-source.
NumberVideoI still sometimes say:
StoryTextthis paper was supposed to be my long weekend plan, but it is so good that I do not think it will even last thursday evening :)
Bold claimImageI've been looking into public benchmarks and am astounded by the poor quality vs. how much we rely on them as an industry.
StoryImageI think Digger's low review score is a reflection of the academic collapse of humanity in general. The movie is a risky achievement in…
StoryImageWe have presents for you! 🎁
StoryImageBig model release today - I wrote about Claude Opus 5.5, GPT-6 Sol, and GPT-6 Luna - plus comparison grids of pelicans by the different…
StoryLinkOne of my favourite things about the new course is the final section:
StatementTextWhat's the best open weight Mixture-of-Experts LLM for coding that fits in less than 60GB of RAM?
QuestionTextAI can write the code.
StatementTextdon't clear that docker image cache fam, its really gone :(
ContrarianImageFolks who are using GLOSSARY.md from /grill-with-docs, how are you finding it?
QuestionTextI feel like I need a word for the type of task that agents are really good at
StoryTextMe acuerdo en la época en la que empezaron los deepfakes allá por 2018 que me preguntaban mucho cómo se podría controlar y darle solución…
StatementTextThe paper on how a team of Gemini 4 agents beat Minecraft is worth a look.
StatementImageThis paper shows why I am skeptical of the idea that frontier models can be easily replaced by cheaper ones (and why you should use the…
Bold claimImageIrrespective of whether software engineering is coming to an end or not, think of it as a roller coaster ride. You are already on it, so…
StatementTextHow would you define spec-driven development?
QuestionTextThis experiment is going horribly wrong 😆
StatementImageI've angel invested in 3 startups shared in the Cursor angel group so far
StoryTextPrompt of the day:
StatementThreadVery useful prompt for Codex with computer use.
StatementImagetech careers are starting earlier and earlier, good thing your kids can listen to this and know how to join the Trillion Dollar Club!
How-toVideoI've never seen an exceptional product that was designed by committee.
StoryText“Sent from AI agent” is the new “Sent from iPhone”
StatementTextAnother blow to Anthropic!
StatementVideoGrok Bot meetups over the next 10 days:
StatementTextThree of the best things to do right now related to personal cybersecurity:
StatementThreadconfession: i've yet to have that moment of feeling that frontier LLMs are geniuses, or super-human at ~anything broadly coherent that i…
StatementTextTo make your agent produce high-quality output, equip it with a way to check its work. Tests are part of that.
StatementImageMuy interesante el testimonio que se recoge en este paper donde se presenta un nuevo benchmark de preguntas teóricas de física.
StatementImageI wrote about the exponential curve of the last 30 months of AI development:
StoryVideoRecursive Agent Harness from Google Research.
StatementImageCan I get someone at Anthropic to bump their cached version of mattpocock/skills?
QuestionLinkeveryone is sleeping on this new JSON extraction model!
StatementVideoWow so Google has just open sourced a tool to automate ANY task on mobile
StatementVideoVivimos en un mundo donde algo que sucede mientras duermes, al despertarte, el vídeo ya ha dado la vuelta al mundo, sí, pero también…
StatementVideoReasoning from scratch, round number 5! This time, talking about log-probability scoring (also a great fundamental concept for loss…
StatementVideoI built a new feature for my blog entirely by voice with Codex Desktop, while I was cooking dinner
StoryLinkPocas mejoras de usabilidad tan útiles como el sistema de recomendación de siguientes prompts que pusieron hace unos días en ChatGPT.
StatementTextI've been doing some hobbyist astronomy with Argon. Here's one of my experiments, where I tasked Argon with reproducing the first black…
StoryImageSecuring all my accounts with AI's help tonight before something bad happens. Also finally stopped being cheap and paid for @1Password
StatementTextReasoning From Scratch: Reinforcement Learning with Verifiable Rewards (RLVR) round 2.
StatementVideoDesde OpenAI indican que han encontrado un dispositivo de grabación en sus oficinas. Han analizado el dispositivo de espionaje usando sus…
StatementImageGuys this is like my superbowl I waited so long for US open weight models to be relevant again. Before today the US models were still…
StatementImageToday we are releasing a new Gemini managed agents version with better caching, lower costs, a Files API, and a Credentials API.
StatementVideoone of the things i really appreciate about muse is this yeti character. it's a perfect mascot for an ai agent: cute, hardworking, sincere.
StatementImagewe’ll tell our kids that the first AI we used was basically a giant scrape of the internet frozen in 2021, couldn’t go online to check…
StoryTextI still have to create a lot of decks for my training biz. The time to create a slide is down 75%
StoryArticleRAG is so back!
StatementImageJ’ai décidé de troqué mes ambitions de carrière et mon salaire de cadre en région parisienne contre une vie tranquille en campagne, du bon…
StatementTextWe are jev-ifying open source models @FeatherlessAI
StoryImageRe-watched Social Network for the first time in years
StatementImageBest way to find out about new @bot features early is to follow more of the team building it:
StatementTextI did not know there was a big AI scene in Kazakhstan. Apparently Higgsfield's entire engineering team is in Kazakhstan and NVIDIA just…
StoryTextNot a lot of positive Kimi 3.1 reactions on the timeline. If you've had time to try the model, what was your impression? (I haven't yet)
StatementTextSo should Terra be the default for Codex now instead of Sol?
QuestionTextMe ha costado 10 años invirtiendo en startups, montando dos de ellas y vendiendo ambas y hablando con cientos de emprendedores el…
StatementTextI built a Command & Conquer style RTS called Battle Drones, mainly using Claude Opus 5.5, which runs in the browser and includes…
StoryImageAs the gap between the internal models at the AI labs and the ones they release grow (due to safety concerns & research advantages)…
StatementTextShared some notes on Python 3.15, using it with uv, and features from Python 3.11 I'm excited to use now that 3.10 is no longer the…
StatementLink🔴 ¡CLAUDE DUET, el ASISTENTE de ANTHROPIC!
StatementVideoNothing kills a literature review faster than this.
Bold claimVideoIts not a new idea, but the lack of giant eccentric monumental architecture in Silicon Valley is a shame. You have lots of weird people…
StatementTextI re-ran an experiment @colin_fraser ran against GPT-4o a while back to see how good it was at adding long numbers, only this time I tried…
StoryImageThe mark of a game changing AI demo isn't how fast it spits out a polished result, but how thoughtfully it responds to an incomplete idea.
StatementTextA malicious release of LiteLLM on PyPI stole passwords and keys from machines, even if your code never used LiteLLM. Here is how it…
Bold claimTextWhen my first AI book was published in 2024, the best AI could do 4 minutes of human work with a 50% chance of success
NumberImage发现一个很屌的输入法,输入中文,自动提示英文,英语学习利器啊!打字的时候,就顺便认识了很多英语单词!
StatementVideoY de repente ahora soy el enemigo de la Casa Blanca 😪
StatementImagePutting myself on payroll was one of the more painful parts of running my one-person S-corp.
StatementVideoWriting good STATA code was very important for my academic career. It looks like that skill is largely becoming obsolete. (Also take this…
StatementImage🔴 ¡GOOGLE LANZA GEMINI 4 ARGON!
StatementImageI see a Tesla topic gaining traction and immediately start thinking about what kind of video I could make.
StoryVideoTop 3 Local AI models
StatementTextDoes anyone else find it lame when founders pick fights with each other publicly?
QuestionVideoToday. The internet breaks.
StatementTextImagine a DeepSeek v4.1 Flash 1017 intelligence and capabilities in under 32GB
StatementTextMultiplayer AI needs to move past the group chat phase. Chats are chaotic, everyone sees everything (creating social & management…
Bold claimTextDeep Blue venció a Kasparov, pero no al ajedrez.
StatementTextMy November course hasn't been shared publicly yet, but it's already ranked top of the search rankings for "software factory course"
StoryTextI gave Claude Code, Codex, and Cursor a self-improving, shared MEMORY.
StoryVideoNew Grok Bot guides added
StatementImageBuilding Hogwarts live in @spawn
StatementLinkDo you need agent skills if you're following a great process?
QuestionTextThis is INSANE...We got an Ultrafast browser agent before GTA 6.
StatementVideoMan, @trq212 always knows what to say.
Bold claimImageEvery so often folks ask me "do you still get nervous doing talks"
Bold claimImageAn absolutely GOATed example of domain knowledge being important
StatementImageMy first open model lecture, covering all the many pieces that build the ecosystem, and the open questions/debates determining where it…
StoryVideoI’ve been playing with Gemini 4 Argon for 7 days — Google gave me early access and sponsored today’s episode.
StoryVideoGoogle Astra nivel esfuerzo máximo. 10 piscinas olímpicos evaporadas para esto:
StatementImageNew Claude Projects* is a nice interface for multi-agent work. It does a good job with having a user-facing coordinator model and…
StatementImageTLDR on the kimi k3.1 pro benchmarks is that it looks like they figured out their RL, where the initial release was a very impressive base…
StatementTextI gave my Dot a $500 budget and 24 hours to plan an entire afternoon for me and a friend.
StoryVideoI’ll hear one song and immediately picture an entire bias edit. Then I open my clips and somehow forget how any of it was supposed to go.
StoryVideobaru tahu hadits ini… di akhir zaman nanti bangsa Eropa / White people bakalan jadi mayoritas populasi dunia? 😲🤯
StatementImageSuddenly Alexandr Wang is all over my X feed.
StatementTextBefore adding another subagent ask what problem it's solving. A bloated system prompt is often the issue.
StatementVideoI fucking feel like having 1:1s with my agents now...
StoryTextAntigravity is now quite insane
StatementVideoThe new Gemini 3.8 TTS models are super-cheap and can generate conversations between multiple voices (from 2,000+, or you can clone your…
NumberLinkCalling all @Bot fans
StatementLinkMy first time seeing unlimited human usage, but limit agent usage. Interesting.
StoryImageThis weird qwen behavior is going to go viral because no one has read traces from mid-tier open models and realized they're almost all…
Bold claimTextVideorespuesta a Yuste a su análisis en Twitch del debate que tuvimos ayer por aquí.
StatementVideoA lot of startups are selling things agents now just do on their own.
StatementTextInference scaling part 1.
StatementVideoSo we already have an open source alternative similar to Jev... and 50x faster?!
StatementVideoFriendly reminder that you can fine tune 500+ open source models in a free Google Colab
Bold claimVideoLa tienda de PC Componentes en Barcelona asaltada:
StatementVideoFinance Twitter is the best
Bold claimTextMETA was up 40% in a month
NumberText“Si al día apagas el móvil 5 minutos, eso hace que el móvil se proteja frente a los virus”
StatementVideoVivix W1 is a real-time interactive video model from @VivixLabs_HQ! 🔥
StatementVideoThe ppl who have been screaming about how higher yields were signaling a fiscal crisis will pivot to falling yields signaling a recession…
Bold claimTextBuild specialized harnesses, folks.
StatementImageThis is the ChatGPT moment for Voice AI.
StatementVideoWhen I was playing with code painting in Argon, I casually asked it to make a realistic scene from pure python.
StoryVideoThe Meta Muse Open License is trash.
StatementTextAnnual S&P 500 returns thru 9/30:
StatementImage
Tweet hooks for AI creators: questions
What is a tweet hook?
The first line of a post: what people see before they decide to read on. On this page a hook is the first line of a real post, up to the first line break or 140 characters, with a link to the full post on X.
Which hooks work best for AI creators?
Among the 407 top posts for AI creators, statement hooks are the most common at 57%, against 64% of all 1,743 posts measured. The "what the data says" table above compares every type. A share, not a guarantee: the author, topic and timing matter too.
How is a top post defined?
A post is "top" when its public view count is in the highest 10% of all original posts we hold, which right now means 27K views or more. Replies, reposts and quotes are left out.
How are hook types assigned?
By fixed text rules, the same for every post: a question ends with a question mark, a list promises a numbered set of things, a how-to promises a method, a contrarian hook pushes against a common belief, a story starts in the first person, a number opens with a figure, and a bold claim uses absolute words. Anything else is a statement.
How often is this page updated?
Posts are refreshed with each data run and this page is rebuilt when the data changes. Last updated 2026-11-02.
Can I remove my post from this page?
Yes. Anyone listed can remove their listing at xleadforge.com/remove; the post leaves this page with it, immediately.
Where this comes from
Hooks are the first lines of public posts by AI creators listed in the X Growth Directory; views are X's public counts. Every example is public, credited to its author and linked to the source on X; we quote only what is needed and never invent a number. Labels come from fixed text rules, not from a model. Anyone listed can remove their listing, which takes their examples off these pages immediately. How we research · Editorial standards