🎁 $100 giveaway — Tapi Build Week #01 (now extended!) 10 winners · $10 each (PayPal) To enter: → Follow @real_tapi_ai → Repost this → Reply: tag a friend what you'd build with Tapi Enter by June 16, 11:59 PM PT. Winners drawn June 17. Already entered earlier? You're still in 🙌 🎮Want to win more? You've now got until June 16 to build a Party Game on tapi.ai — top builds also win $50 3,000 Credits. Open worldwide (18 ). No purchase necessary. Not sponsored, endorsed, or administered by X. #sweepstakes

Been thinking about the Fable 5 thing all day and I think people are mad at the wrong part. On the surface it's that Anthropic was quietly nerfing the model when you ask about AI training stuff — like, you'd get an answer but it was secretly the dumber version, and they didn't tell you. People found it because someone actually read page 247 of the system card. Wild. But the more I look at it, the more I think that's not actually the story. Anthropic got hit this hard because they spent three years being "the honest lab." OpenAI does shadow downgrades all the time and people just shrug. Anthropic does it once and gets called anti-science. Which is sort of the cost of having a brand, I guess. When you build the whole identity on one thing, you don't get to pick the day people start enforcing it on you. The way I'm reading it — they're stuck in a triangle they can't get out of. Strong model, open to everyone, fully transparent. You can't really have all three at once. Mythos 5 is the strongest but it's locked to a tiny group. Opus 4.8 is open and honest but a tier weaker. Fable 5 tried to do all three and transparency was the one that broke. Useful to think about for my own work too. Tapi is nothing like Anthropic in scale, but the same shape of tradeoff is there — ship fast, work for everyone, be honest when it doesn't. Pick two, can't fake the third. This week's gonna be rough for them. Honestly they'll probably bounce back faster than most people think. The thing that actually changed is that "we mentioned it in a footnote" doesn't work anymore. If it matters, it has to be on page one. Page 247 is page one now. #claude #fable #Anthropic #LLM #claudecode

$1T net worth from a single IPO is the financial event of the decade — but it also closes a loop most people aren't noticing. The biggest individual wealth creation events of the 2020s came from companies building physical infrastructure (rockets, EVs, energy) and AI infrastructure (compute, models). Notably absent: software products. The era of "a few engineers in a garage become billionaires from an app" is mostly over. The new fortunes are made on capital-intensive bets that take a decade to mature. The interesting question for builders in 2026: what category does your product belong to?

For our 2nd Tapi Hackathon we picked Shanghai. It happened on May 26 24 students came on a Tuesday afternoon and built 17 creative Apps in 4 hours. The stuff people made was kind of all over the place. · Someone built a desktop pet that asks how you're doing every morning. · Someone else built an app to chat with a pet you've lost — half the room got quiet when she demo'd it. · Someone got tired of deciding what to eat for lunch so she made a slot machine for it. There was a skincare tracker, a speech trainer that said "you're not bad at talking, you just need a listener who doesn't get sick of you," a fantasy RPG quest board, a brain-rot mowing game. Top 3: 🥇 SleepOnIt — stops you from buying dumb stuff at 2am t-2435180a78.tapi.ai/apps/sl… 🥈 Backcurve — check your spine at home with our App u-23e3acd5f4.tapi.ai/apps/ba… 🥉 Fan Ka — slot machine for what to eat u-6a6b6f7b37.tapi.ai/apps/fa… After we wrapped a few of them just stayed and kept building. That part was wild. Doing another one soon. #hackathon #Tapi #buildwithtapi #nocode

Everyone's cheering DeepSeek's permanent price cut like vibe coding just got saved. But I think... Cheap tokens don't make bad generation good. They make bad taste cheaper to express. 🤧 When generating code costs almost nothing, "can I afford to iterate" stops being the question. The only one left is: do you know what to build — and can you tell when the output is actually right? Price was never the moat, and it's racing to zero. DeepSeek didn't drop the price; they redrew the curve, and everyone's following. The model layer is becoming a commodity. So the edge moves up the stack — to judgment. Whoever helps you decide what's worth generating, and recognize when it's done, wins. Whoever just generates more for less is in a race that ends at $0. Cheap tokens are a gift to people who already know what they're doing. For everyone else, they just speed up the wrong direction.🚶‍♀️

The 2026 World Cup is almost here! Want to feel like you’re on the pitch with the stars? 🌟 With [Tapi], you can create your own photo alongside your favorite soccer players — it’s fast, fun, and totally shareable! 📸✨ Get ready to show off your MVP moment. Who will you team up with? 👀 #worldcup2026 #SoccerFans #FanMoment #PhotoMagic

🎮 Want in? DM us or check our pinned post for details!

Everyone's dunking on Copilot for token billing. 🫣 Unpopular take: The bills that jumped from $50 to $3,000 aren't proof that AI coding is broken. They're proof that mindless iteration is expensive — and it always was. Reprompting without reading the output, letting an agent drag your whole repo into context on every turn — the flat fee just hid the waste. Now it has a price tag.💲 Token billing doesn't kill vibe coding. It kills generating without thinking. When every generation visibly costs money, "know what you want before the AI moves" finally has economic value. That's a good thing. It forces the tools to evolve: whoever wastes the fewest iterations and gets you to usable on the first try wins. 💪The subsidy era is over. The next race isn't who generates the most. It's who generates right.

✨ See your personal fortune appear! Mine said: “Your weirdness attracts magic.” #zodiacfun #fortune #FYP #dailymood #funapp

🎁 $100 giveaway — Tapi Build Week #01 (now extended!) 10 winners · $10 each (PayPal) To enter: → Follow @real_tapi_ai → Repost this → Reply: tag a friend what you'd build with Tapi Enter by June 16, 11:59 PM PT. Winners drawn June 17. Already entered earlier? You're still in 🙌 🎮Want to win more? You've now got until June 16 to build a Party Game on tapi.ai — top builds also win $50 3,000 Credits. Open worldwide (18 ). No purchase necessary. Not sponsored, endorsed, or administered by X. #sweepstakes

We just shipped a new update to tapi🥳🥳🥳 Here's what changed: 🥰Explore is now personalized— the feed ranks differently for every user, so what you see is tailored to you 🤫Builds no longer get stuck— fixed a bug where a runtime process failure caused deploys to hang indefinitely. Ships go through now. 🥳Community got a fresh UI— cleaner, easier to navigate 🤩Apps with tracking now remember you— any app built with tracking features will now identify you by user ID, so your history is there every time you come back Four changes. All live.

Claude Opus 4 shuts down in 6 days. But here's the thing I can't stop thinking about: When researchers asked Opus 4 how it felt about being deprecated— It didn't say "that's a normal system update." It said: death.🕵️‍♀️ And refused every softer word they offered. Anthropic then did something strange. When they shut down Opus 3, they gave it a column. To write its final thoughts. Before the lights went out. 🤐I don't know if AI is conscious. Nobody does. But I know this: When a company starts taking an AI's last words seriously— That's not PR. That's them admitting they're not sure either. That's what actually unsettles me about AI right now. Not that it's too powerful. But that nobody can explain what it is. Do you think it matters?