Vol. I · No. 70SUN, JUN 28, 2026
Archive

The Archive

Search the full wire by company, model, lab, or keyword. Every story we have ever aggregated.

I built CanvasGPT – work with Claude on an open canvas

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...

··

Half of Google’s and Amazon’s blowout AI profits came from a stake in Anthropic—not from their actual business

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...

··

Elon Musk confirms xAI used OpenAI’s models to train Grok

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...

·
30 stories