🤖

Be sure to create your FRESH Polymarket account via THIS link: polymarket.com/?via=alphastac… --- very important that everyone does this

who's the first person that believed in you more than you believed in yourself?

@CoinQuantX nice video!

A profitable strategy on paper can fail in live markets because of transaction costs. Fees, spreads, and slippage change everything. #trading #goldtrading #daytrading#algorithmictrading

Hey guys, if you're in Ho Chi Minh City, Vietnam you can come celebrate Bitcoin Pizza Day with me & the gang! Sign up here; luma.com/s4f2t1pd Limited spots. Talk crypto. Talk AI. Enjoy some pizza. Good times. See you there 🤝

@rsa @TIDAL Boards of Canada ~ love New stuff just dropped but dropping this here for newbz youtube.com/watch?v=Ng2vTX…

Rolling out better download functionality for @tidal. Way more reliable, and much faster. Look out for it soon.

@LimitlessFT LLMs => Agents => Robots

A $25,000 investment in Sandisk 1 year ago would be worth $1,000,000 today. That's a 40,000% return. We just found the next biggest AI trade... and the secret lies in the AI infrastructure bottleneck: -> INTEL and AMD are up 3-4X this year because CPUs unlock agentic AI -> AI memory stocks $MU are up 700% because AI needs to remember you The next major bottleneck is obvious but everyone's missing it, we unpacked it all in today's episode 👇

Justin Sun @justinsuntron just launched @BAI_AGI -- a unified AI gateway that gives you ONE API key for Claude, Gemini, DeepSeek, Kimi, and more. NO KYC. PAY USDT! I built an app to put it to the test head-to-head against OpenAI and Google. Enjoy! youtube.com/watch?v=8OBfz4…

@TylerDurden @skittlewood lol

@skittlewood What in the photo is Vietnamese?

Vietnamese > Thai food Change my mind

@AshConnell @threejs the frame rate starts to topple and the leads to an eventual crash, if I don't hard stop the server okay I take your point, so fare I've been focused on just getting it working and the game loop to be fun ~ but measuring is definitely the logical next step for optimizations

yeah but what issues? low frame rate? crashing? set up a stats overlay to show draw calls, cpu and gpu milliseconds, number of triangles etc. the solution is always to measure and find out what is slowing it down, then the fix is usually more obvious. the problem is that it could be a number of different things, so you have find out which one.

a web based @threejs streamed open-world with 100 unique players running on my 6 year old iphone 12 pro. ai lets me fail fast and go in a bunch of directions until something lands. this would have taken me years without ai! i ended up building a budgeting system for both plots (lods) and avatars (ghost+cull). vram on mobile devices is hyper critical - safari will just kill the page if you go over ~150MB. and yes... every player here is representative of a real player with completely unique wearable meshes each, although animations are all shared and happen on the gpu.

@ThirdTimeIan @threejs @OpenAI ah, okay I tried to avoid blender.. and perhaps that's why everything that's built is shite; looks bad and is super glitchy so -- reduce poly count -- combine meshes and mats .. any other good tips?

@alphastack1 @threejs @OpenAI The main helper was Codex / GPT writing scripts to a) reduce poly counts and b) combine meshes and materials - it did it all via Blender (headless or CLI I think - I never even saw it open the app or anything)

Well I think we did pretty good @threejs @OpenAI What would have never run on web (and was even a struggle on Unity) now runs pretty damn solid, 20 unique horses + jockeys down to 538k tris, 44 materials and 141 draw calls (think I can cut that down 30% still too)

@AshConnell @threejs I am trying to build a fully procGen 3d environment, with different heights and angles etc - and fill it full of mobs (like 100-300) that swarm at the main player. I have had issues with the number of mobs mostly, and also with the system keeping track of the player-on-terrain

@alphastack1 @threejs what performance issues are you running into?

when Claude Opus 6 tells you to "stop spiraling and go to bed" 😵‍💫

Bitcoin is getting hard forked into a new chain called eCash -- and every BTC holder will receive a 1:1 airdrop. This hard fork will create a new main L1 and 7 new sidechains at the same time: prediction markets, a DEX, NFTs, identity, quantum resistance, privacy, and one ??? coming soon slot. youtube.com/watch?v=buhJpD… @Truthcoin @LayerTwoLabs

Thanks to all my Youtube subscribers! Just hit top 25% of creators with 1K–5K subs -- 38,923 monthly views

@perea_ai @techopsasia Build Stuff meetups are fun

@techopsasia hi! also living in saigon now, do you know any events or communities that may be interesting? working on agentic AI / autonomous agents / etc will be nice to know more people doing the same

I keep coming back to Vietnam. There's something happening here that's hard to explain until you're on the ground. In Saigon 🇻🇳 right now, delivering AI and cloud projects. Our clients are global and increasingly they don't care where the work comes from, they care what comes out the other end. That shift, from outsourcing to outcomes, is exactly why Silicon Valley is planting roots here. TinyFish AI didn't come to Vietnam for cheap code. They came for brilliant engineers. Worth paying attention to.

@Pizza_DAO @standwithcrypto @brave @WrldPizzaChamps 🍕🍕🍕🍕 #SAIGON VIETNAM! Luma page coming soon @mantle_vn @Jul_K_Lee @gem_decrypto

🗺️🍕🥳 Bitcoin Pizza Day is in 3 weeks! Find your city's party at globalpizza.party. 🔍

Adobe just dropped their Creativity Connector inside Claude - meaning you can now run Photoshop, Lightroom, Premiere, Illustrator, and Firefly tools directly from a Claude chat. youtube.com/watch?v=Avl5Q-…

@RetroChainer Google's TimesFM does the same

I told Claude to predict the next BTC candle using Kronos. didn't touch anything. +$4,200 overnight. Kronos isn't another LLM wrapper. it's a GPT-style model trained on 12 billion candles from 45 exchanges. Binance, NYSE, NASDAQ, LSE. it reads OHLC the way GPT reads English. > words = candle patterns > sentences = OHLC sequences > the corpus = every trade ever printed > the edge is simple: P(next | history) = softmax(W · h_t) same math GPT uses to predict the next word, except the "word" is the next candle. no twitter sentiment. no blog scraping. no hallucinated confidence. just statistics over 12B real price moves. I plugged Kronos into a 5-min BTC Polymarket bot. input: last 50 candles output: probability next candle closes up trades only when |p_model - p_market| > 0.04 72 hours: $300 → $4,500 ChatGPT looks at a chart and pattern-matches against an article it read. Kronos looks at the chart and remembers 12 billion of them. 87% of Polymarket is exit liquidity. one of these models is actually trading the market. github.com/shiyu-coder/Kr… 21.8k stars, free, MIT license. Tsinghua University release.

TAKE PROFITS when you're up .. because no one STOPS on a winning streak