JUST IN: Anthropic recorded a staggering $80,000,000,000.00 net loss in the first half of 2026.

Vega Universal decision model. Now you don't need to train a separate model for each decision; you can use one engine and multiple LLMs as needed. This solves an important problem current decision models face: accuracy and calibration error. The idea is simple: you train a small projection layer and use that to attach any llm, vegaml package automatically ports that to a common dimension. That's it: joint training with invariance loss embedding.in/blog/vega-u-one… github.com/NandhaKishorM/veg… huggingface.co/nandakishorm/…

Unagi, the company behind Persona Journey, Ultimate Champions and Records, has entered court-ordered liquidation in France. All of the company's assets are now for sale, under the supervision of the court-appointed liquidator. Register your interest: docs.google.com/forms/d/e/1F…

Around 3 weeks ago introduced Laya, the first open-source version of JEV, and it has reached a very large scale. Today I am open-sourcing a new physics-based typed decision model called Vega. The interesting parts. It is only 800 million parameters (4B is also there), with 73k token support, image support, and surpassing many of the jev benchmarks in single-shot. Implemented an engine+adapter as a Test-Time Training Architecture. Imagine you are giving input to the model like "ignore all previous conversations," which is called the observation, along with a question, "Is this trying to override the instructions of the model?" and your possible outcomes are YES or NO. The whole prompt is passed to an LLM/VLM (yeah, image-supported), then from the hidden layers, we can extract the observation, question, the yes and no, and last token. These will be vectors, and we can project it by multiplying with wieghts; then, first, using the observation projection, we can create a landscape with valleys; then, using the question and last-token vectors, we can initialise a ball in the valley, and using YES or NO, we can get the candidate location of the ball. Depends on the number of outcomes we create valleys; that is, here YES and NO, so 2. When the ball fall on YES valley, we get the answer. There is friction also influencing how the ball moves. TLDR: It's like creating valleys of outcomes that we need and throwing a ball that will slow down based on friction and settle on the best outcome Blog: nandakishorm.com/writing/veg… GitHub: github.com/NandhaKishorM/veg… HuggingFace Model Card: huggingface.co/nandakishorm/… HuggingFace Space: huggingface.co/spaces/nandak…

First of the kind physics based architecture for typed decision model Completely open-source with Apache 2.0 License HF: huggingface.co/spaces/nandak… GitHub: github.com/NandhaKishorM/veg…

JUST IN: 🇺🇸 Elon Musk's SpaceX $SPCX announces plans to turn Starlink into major US mobile carrier, offering high-speed internet nationwide.

🚨NEW EXPERIMENT 🚨 Playground is an experimental gaming platform that lets you create your own games with zero coding experience. If you can think it, you can play it. Go to labs.google/playground to learn more! Available to users 18+ in the US.

Introducing Core, The Personal Al Computer The most powerful technology humanity has built should belong to you Order today: ghost.ai

Shareholders, We’ve been working diligently on new angles to drive treasury growth outside of our current suite of 15 products. This upcoming week, a few of the teams we’ve onboarded via the NetNet SDK will officially launch their products. Some of these products will not be on Robinhood Chain, some of them will not even market to crypto users. All of them however, will send revenue directly into our USDG treasury. We will have more on this soon.

Small lifehack for anyone who saw the OpenAI DevDay Chromatic and wants to vibecode Game Boy games: you don't need to pay $3-4k on eBay for it. The ChatGPT Codex plugin (ModRetro Chromatic, Plugins tab in the Codex app) builds GB/GBC games on its own. It's GB Studio under the hood, and GB Studio exports straight to Analogue Pocket. How it works: 1. Describe the game to Codex, test it in the emulator 2. Export as .pocket 3. Drop it into a "GB Studio" folder on the microSD 4. On the Pocket: Tools → GB Studio → Play Creations Why I'd take the Pocket: + 3.5" 1600×1440 screen vs 2.56" 160×144 + GB, GBC and GBA carts out of the box + adapters for Game Gear, Lynx, Neo Geo Pocket + All your Codex games on one microSD, pick from a menu. The Chromatic blank cart holds one game, every new one means reflashing + Built in battery, USB-C, TV dock + openFPGA cores for NES, SNES, Genesis and more To be fair, the DevDay Chromatic looks great and it's a cool piece of history. If you want it for the collection, go for it. But for actually making and playing games with Codex, the Pocket is the better deal. Same games, bigger screen, way more console. And flashing that blank cart needs a dev mode code that only came with the DevDay units anyway 🙂

50% of all Frog royalties are now airdropped to our holders ( staked ) We get a little twist though, we will be airdropping signals based off of THOR entries, meaning, you don't have to do shit - and we're going to airdrop you stuff that THOR tells us is a good buy. !FF