just built a TUI that trades LTC/USD it uses HTT + Jev AI (laya) as final filtering those decision/classifier models fail bad at math stuff coupling them with some tailored indicators could make some difference (?) 100% out-sample with 0.2% trading fee

A store's existing CCTV turned into a shelf-attention heatmap. No new hardware. The digital twin is scanned with a phone, deploys in under a day. This is @CactusXR. Running at a 5,000 m² coop in sweden: 6% basket-attachment lift, 45 minutes a day saved per associate. Every retailer on earth already owns the cameras. That's the whole insight.

Bot detection & human verification will be one of the most urgent demands for businesses over the coming years. Agent swarms will suffocate every website and form; small companies and government websites are most vulnerable. There is a huge gap in the market for this right now. When we looked at what offerings were in the market to use at X, there was not a single company that brought together all the latest technologies so we had to do it all in-house.

Welcome to the future of retail, where AI tracks what catches your attention Soon, everything we do in public will be analysed and optimised for more sales Writer: Ian

implemented 𝚒𝚜𝙿𝚛𝚒𝚖𝚎, but using @typesafeai Jev + @aisdk made a prime spiral of first 10.000 numbers it says true most of the time lol

Introducing 𝚒𝚜𝙴𝚟𝚎𝚗 / 𝚒𝚜𝙾𝚍𝚍, but with @typesafeai Jev + @aisdk Now available on NPM: 𝚗𝚙𝚖 𝚒 𝚓𝚎𝚟-𝚎𝚟𝚎𝚗-𝚘𝚍𝚍

Jev may have its (narrow) use cases but the hype you see around is the perfect representation of the fact the greatest part of the AI bubble don't know what is important and what is not. Jev is a minor thing happening on AI compared to all the rest, yet the hype exploded.

Introducing fastbrowse. An open-source fast browser agent. Jev picks each action directly from the page, an LLM plans and every claim cites an exact quote. It's significantly cheaper and faster for agents to browse the web now. Early and experimental, but very promising. fastbrowse.ai

Here’s a detailed thread about how we built this, the tools that were used, and how we tried to keep user and business data privacy at the core of the system 🧵

⚡️DID YOU KNOW: Bitcoin's source code is buried in the Arctic World Archive, a vault designed to survive for 1,000 years. In 2020, GitHub archived Bitcoin's entire codebase onto 186 reels of archival film and deposited them in a decommissioned coal mine beneath the permafrost in Svalbard, Norway. The archive holds 21 terabytes of open-source code including Bitcoin, Linux and Python.

Breaking: Browser Use + Jev = Ultrafast ⚡ Findings flights took 7s and cost only $0.0039 🤯 > new action space every step > DOM state space > small LLM fallback to type (this video is at 1x speed btw) Built a tiny open source browser agent. try it below ↓

Workflows but shadermaxxing.

what's that ONE book any soft eng MUST read? (no Clean Code pls)

Funding onchain teams is hard because the failure rate is extremely high. There are teams that are great on paper but do horribly when building onchain products. Sometimes it’s a mismatch with the user base and culture, but more importantly, onchain products have to get a lot right. If they don’t, things can implode and people can lose a lot. Or they get lucky, become successful and start printing revenue quickly. Then ego gets tested, and many times people fail that test. In many ways, building onchain requires a very specific understanding of culture and onchain mechanics, but most importantly, you have to check your ego and stay humble if you start to win. Very different from non onchain startups where it takes years to get to $100m in ARR and the founders have had to grow along the way and earn the right. With onchain, it can happen in a few months and it takes a very specific founder to understand this responsibility. Which is why very few win long term in this space.

Bitcoin is going to $1 million without Clarity. The government can go f*ck itself.

(っ◔◡◔)っ₿

We are investing $8,000,000 into 20 teams. Some founders come from SpaceX, Stripe or Google DeepMind. Others come with an idea and no logo. We back both.

So happy that GraphQL is not a thing anymore. It was awful.

What’s the best way to e2e test APIs on @kubernetesio microservices, locally, mocking only third-party http requests? I’ve heard of kind, but curious to know if there’s other/better ways to do this.

stupid amount of value in chinese github repos