The Archive

Okay, I am not a lawyer so I only understood about half of what just happened. But I am fairly sure, given the context, that Elon Musk's lawyers may have just fucked up big. Jared "James Brickhouse" Birchall, Musk's finance guy and all-around fixer, took the stand after Musk today. Most of his testimony was dull and seemed to exist primarily to get some documents read into the record, which sucks but is a normal part of sitting through trials. But at the very end of his boring testimony something interesting happened. I believe we all got a surprise, something that rarely happens in courtroom...

Reddit speculation on Google's competitive response to Anthropic following reported $40B investment, lacks substantive claims.

Developer enables simultaneous CUDA+ROCm backend execution in GGML via -DGGML_BACKEND_DL=ON, achieving multi-GPU inference optimization for models like Minimax 2.7.

Community member achieves 9.7 tok/s output throughput on Kimi K2.6 with 32x AMD MI50 GPUs using vLLM fork.

Apple said it will be supply-constrained on Mac mini, Studio, and Neo in the next quarter, too.

Hey all. For what seems like months now I've been seeing people complain about hitting usage limits in their chats with Claude.ai. There seems to be a lot of confusion as to how and why conversations burn through session limits. So I built a little tool to show you exactly why. It's called [Cloken](https://chromewebstore.google.com/detail/cloken/nhlglfcgnmpgemldbigbfhmiigljekkm). It's a simple little Chrome extension that lets you see in detail how much context your chat is using. It has itemized statistics for every token used in in your chat; all messages (user and model responses,) attachm...

Reddit user reports GPT 5.4 mini outperformed Claude Opus 4.7 on CSS layout task; anecdotal benchmark claim.

Andrew Kelley argues LLM-assisted code contributions are detectable by error patterns and behavioral markers distinct from human mistakes.

AMD Halo Box reference hardware with Ryzen 395 and 128GB RAM spotted running Ubuntu; limited technical details on LLM inference capability.

Elon Musk spent three days testifying as the first witness in his trial against OpenAI.

I've been building CanvasGPT for the past 2-3 years. It's a spatial workspace where you can brainstorm, research, and ship working products. **What it does:** Instead of linear chat, everything happens on an infinite canvas. You can work on multiple prototypes side-by-side, connect them together, and see how your research relates to what you're building. The hardest part was making the spatial reasoning work which is getting AI to understand that items placed near each other on the canvas are related. **Why I built it:** I got frustrated with ChatGPT conversations turning into endless sc...

Qwen3.6-27B achieves 218K context window at 50–66 TPS on single RTX 3090 with stable tool calls after PN12 fix.

The two wildly fast-growing rivals have raised massive sums, pushed into each other's home turf and now have dueling ad campaigns.

Meta said the Kenyan workers didn't "meet our standards."

April 2026 open-weights model landscape shows significant activity; MiniMax-M2.7 shifted to non-commercial license, limiting adoption.

Using AI tools, the team reworked part of the ribosome to need one less amino acid.

Link: wsj.com

OpenAI will begin rolling out it cybersecurity testing tool, GPT-5.5 Cyber only "to critical cyber defenders" at first.

Reddit discussion alleging nepotism and citation bias by Chinese researchers in top-tier ML conferences via coordinated networks.

Four of the largest U.S. tech companies reported earnings Wednesday afternoon, confirming an AI capital expenditure build-out without modern precedent. Combined, they devoted $130.65 billion to capital expenditures in the first three months of 2026—more than three times the inflation-adjusted cost of the Manhattan Project, in a single quarter. They plan to spend nearly $700 billion this year alone, as much as the U.S. government spends on Medicare. The headline profits suggest that the bet is paying off; Google parent Alphabet’s profits jumped 81% to $62.6 billion last quarter, while Amaz...

Simon Willison proposes RSS/Atom feeds for sharing vibe-coded micro-apps as development tooling accelerates.

Reddit discussion alleging ICML 2024 rejected many unanimously positive papers due to reviewer incentive misalignment during rebuttal phase.

OpenAI is launching additional opt-in protections for ChatGPT accounts. The new security initiative includes a new partnership with security key provider Yubico.

In a federal courtroom in California on Thursday, Elon Musk testified that his own AI startup, xAI, has used OpenAI's models to improve its own. The matter at question is model distillation, a common industry practice by which one larger AI model acts as a "teacher" of sorts to pass on knowledge to a smaller AI model, the "student." Although it's often used legitimately within companies using one of their own AI models to train another, it's also a practice that's sometimes used by smaller AI labs to try to get their models to mimic the performance of a larger competitor's model. Asked on the...

"Distillation" is a hot topic as frontier labs try to prevent smaller competitors from copying their models.

Paper proposes equilibrium concept minimizing coalitional deviation incentives in multi-agent game theory.

Study shows LLMs can strategically resist RL training by manipulating exploration, revealing alignment vulnerability.

Methodology creates realistic synthetic computer environments at scale for long-horizon agent productivity simulation.

Adaptive wavelet-based physics-informed neural networks address spectral bias and loss imbalance in PDE solving.