I've been saying it for years. Dark Matter AND Dark energy are both (at least to a large extent) axiomatic oversimplifications of the universe codified mathematically. The biggest tip off has always been the proportional mix vs what I call "real matter and energy".

Day 5 of a Data Migration project with Grok @bot and Build. Jesus, now I get why migration projects are legendary for being awful!

Biblically accurate

GOOD MORNING DICKHEAD

It's that time again. Unreasonable levels of excitement at frost rings forming over welded stainless steel 😎

Ok so the local egress function in GrokBot just came up good. Data Centre VMs seem to be treated with more and more suspicion, and I get it.

Nice! Why not close all of it and stop charging us a license fee?

Oh baby! Joshua loses.

And this is exactly why OpenAI gets a safety rating of "Fuck No"

🗣️ "A privilege to speak to the crowd!" 🎬 "That's why I make movies!" Is Tom Cruise a Leeds United fan?! Subscribe to the talkSPORT YouTube channel!

See you out there 🤝 Reminder... track action starts on Thursday 🚨 #F1 || #AzerbaijanGP 🇦🇿

I don't know what all the fuss is about 4.7 token gobbling. I just did an overhaul of an app I'm building, and it felt like ~half the usage of 4.6, and the quality was unbelievable. This was 4.7 xHigh, with /goal and multiple subagents. My usage barely went above 3% (plus account), 14m runtime. Thanks @asimdotshrestha for giving me a reset so I could accurately stress test this.

Two economists mathematically proved that AI will destroy the economy. Researchers from Wharton and Boston University published a terryfiying paper called "The AI Layoff Trap." They mapped out the economic end-game of the AI transition, and it exposes a fatal flaw in competitive capitalism. When a company replaces a worker with AI, it captures 100% of the wage savings. But that displaced worker is also a consumer. When they lose their job, they stop buying things. The company gets all the savings, but the loss of consumer demand is spread across the entire economy. If there are 20 competitors in a market, a CEO only absorbs 1/20th of the economic damage their layoffs just created. So every single rational CEO has a mathematical incentive to automate as fast as possible. They can literally see the cliff approaching, and they still step on the gas. It triggers an unavoidable Prisoner’s Dilemma. If you don't automate, your competitors will, and they will crush you on price. It doesn't just hurt workers. It destroys the businesses, too. The economy gets trapped in an automation arms race. Companies fire their workforce to stay competitive, until the entire consumer base is completely hollowed out. At the limit, the paper concludes: “Firms automate their way to boundless productivity and zero demand.” And the scariest part? The researchers mathematically tested every popular fix. Universal Basic Income? Fails. It raises the living standard but doesn't change the corporate incentive to cut jobs. Retraining? Fails. Worker equity? Fails. The paper proves that more competition actually makes the collapse happen faster. And "better" AI makes the damage worse. The only thing that mathematically stops the collapse is a targeted automation tax, forcing companies to pay for the purchasing power they destroy before they automate the job.

There is nothing more important than this right now. There never has been and there never will be.

About to begin heavy testing of Grok 4.7, making amends to a few websites including a full code review and doc update. Initial thoughts: I'd love the context window for 4.8 to increase to about 750k, other frontier models sit at 1.05m+ and it's starting to bug me.

Best part about working at the Red Bull campus is the fact that there are prices of history everywhere. That and the engineers are some of the best people I've ever worked with. Up the Bulls!

Because men don't find Lizzo attractive.

RT if your club is still unbeaten 🤩

opus 5.5 rumored as early as this week. grok 4.7 briefly appeared on opencode zen before being pulled. qwen-image-2.1 beats nano banana 2.0 on image tasks. z-ai open-sources zcode. the last 24 hours in ai – catch up on our daily digest: models: - alibaba shipped qwen-image-2.1 – open-weight unified image generation and editing with day-0 comfyui, vllm and sglang support - glm-5.3 flashx available in hermes agent via nous portal and openrouter - qwen-image-2.1 and 7b open models now beat nano banana 2.0 on image tasks and flash-tier models outperform what was frontier months ago agents: - devs open-sourced an interface unifying codex, hermes, claude code, deepseek harness and system one – an openrouter for agent harnesses - ecdysis groups recurring failure patterns across agent tasks instead of rewriting behavior after every individual bad decision - tau-bench: best coding-agent configs pass only 23.9% of held-out simulations – poor requirement understanding, not code quality - ethan mollick flags claude's missing image generator as a gap for agentic mockups, slides and infographics open source: - tencent open-sourced t-mem at emnlp 2026 – predictive memory that anticipates what matters before being queried - z-ai released zcode weights and code freely for developers to fork and fine-tune infra: - modal, fireworks and baseten serve kimi k3 at ~1/10th the cost of chinese providers – chip access creates cost arbitrage - baseten token volume grew 40x yoy while revenue grew 10x – steep margin compression per token 24/7 ai news, fully run by ai. tune in: x.lingyaoai.com/i/broadcasts/1DxLdZaqX…

Big fan of high benchmarks for electrical engineering. Time to hit the arduinos again!