Tuesday, August 4, 2026 · about a 2 minute read
Today's news keeps circling the same quiet tension: AI systems are getting more capable, but the interesting action is in making them lighter, more honest, and useful to more people.
Get the calm version of AI news.
One email a day on what is actually happening in AI, in plain English. No hype, no doom.
Free. One calm email a day. No hype, no doom.
A developer got an 80-billion- model running on a Mac and a 35-billion-parameter one on an iPhone, using aggressive compression. If this keeps working, the question 'can I run a serious AI locally without paying for a cloud service' gets a real answer much sooner than expected.
A well-upvoted essay argues that AI tools reward people who already know their domain deeply, because they can spot bad output, ask precise questions, and steer the model usefully. If you are a beginner handing the wheel to AI, you may not even know when it has driven you off the road.
Researchers tested sixteen different language models on open-ended questions and found the combined output had the diversity of roughly 1.7 distinct voices. That matters if you are using multiple AI tools to get a 'second opinion,' because you may really just be asking the same person twice.
Simon Willison flagged a new term worth knowing: 'meat proxy,' meaning a person who just copies AI output and passes it along without reading it. It is a funny phrase for a real problem, and if your job involves reviewing or communicating anything, it describes a failure mode you have probably already seen.
A new tested large language models on Tibetan medicine and found they perform much better on Western medical traditions than traditional ones. If you or someone you care about relies on non-Western medical knowledge, the AI assistant you ask may simply not know that world very well.
Get this every morning.
RAG (retrieval-augmented generation) is how AI tools look things up before answering, like a researcher pulling files before writing a report. This paper shows that checking each file one at a time fails for complex questions that require connecting two or more sources together, the way a good answer about a legal case might need three separate documents to add up. Understanding this helps you know why an AI tool can confidently give you a wrong answer even when the right information was technically available to it.
Researchers found that a small, specific cluster of neurons inside a model can reliably predict when it is about to misuse a tool, like calling an API it should not or skipping one it should. This is early work, but the idea that misbehavior leaves a detectable fingerprint inside the model is the kind of thing that could eventually make AI much safer to deploy in real workflows.
That's today. See you tomorrow.
Get this every morning.
One email a day on what is actually happening in AI, in plain English. No hype, no doom.
Free. One calm email a day. No hype, no doom.
The book behind this newsletter
How ChatGPT, Claude, and Modern AI Actually Work
The trick is small. The world it built is not.