8 posts tagged with "ai-news"

Almost every "secure your LLM app" guide gives the same advice: before a prompt reaches the model, strip the personal data out of it — swap names, emails, and account numbers for [REDACTED] or <PERSON>, then call the model.

It sounds obviously right. I assumed it was, too. Then I went and read the research on what masking actually does to a model, and the papers point the other way. The short version: mask your logs, not your prompts.

Someone posted spec-driven development on LinkedIn this week as the answer to vibe coding — to prompting an agent, half-understanding what you're building, and ending up with code you can't vouch for. The linked toolkit has 127,000 stars and comes from GitHub itself. The pitch lands.

So I installed it and pointed it at a deliberately trivial task. One of the three principles it wrote for me was a dependency policy I never asked for — hold that thought.

Twenty minutes isn't a verdict, though. Two engineers have tested this properly, on real problems, long enough for the seams to show. They used different tools, on different continents, seven months apart — and both reached for the same comparison, unprompted: the last time our industry tried to generate working code from documents. On the one question that decides whether any of this survives contact with AI features, they flatly contradict each other. Neither has a measurement.

Most "run it locally" model announcements come with an asterisk — smaller, weaker, a toy version of the real thing. Meta's newest release doesn't: Muse Glimmer, a 30B multimodal model built specifically for agentic work, fits on a single consumer GPU and beats larger models on the benchmarks that actually measure agent behavior.

In June 2026, the AI world got a new buzzword: loop engineering — roughly, disciplined design of a single agent's tool-calling loop. Six weeks later it was supposedly replaced by graph engineering — wiring up several agents at once — killed by twelve words that 3.1 million people saw:

Don't know what loop engineering is? Don't worry — neither did most of the people declaring it dead. Here are both terms, how a name became an obituary in six weeks, and the twist nobody checked before writing about it.

Protocols · AI News

The Model Context Protocol just had its biggest revision since launch. The 2026-07-28 specification doesn't add a feature — it rewrites how every MCP server talks to every client. If you've deployed an MCP server anywhere past "runs on my laptop," this one touches your infrastructure, not just your changelog.

OpenAI shipped GPT-5.6 on July 9 — a family of three models (Luna, Terra, Sol) — and the headline claim isn't a leaderboard score. It's an efficiency number: frontier coding performance on less than half the output tokens. If you build agents, that's a claim about your bill, not about bragging rights. It's also exactly the kind of claim you should measure yourself.

Look at almost any current model leaderboard and the top is a wall of Anthropic and OpenAI. Then, sitting in the top 10, there's one outlier that isn't proprietary at all: GLM-5.2 from Z.ai — open weights, MIT-licensed. That's the story worth paying attention to.

Infrastructure · AI News

The AI ecosystem is quietly going through the same transition web infrastructure went through years ago: the performance-critical pieces are moving off interpreted runtimes onto systems languages like Rust. LiteLLM rewriting its AI gateway in Rust is the clearest evidence yet — and a sign the AI stack is maturing from experiment into production infrastructure.