Llama

Around March, Google told Meta it couldn't supply the full Gemini capacity Meta had tried to purchase, delaying some of Meta's internal AI work. Meta then asked staff to be more efficient with tokens. Other Google Cloud clients were squeezed too.

The thread's flip: this isn't Google restricting a rival's access, it's a compute shortage dressed as a restriction. And the buried oddity is Meta, which builds Llama, leaning on a competitor's model in the first place.

Wayfinder sits behind any OpenAI-compatible client and decides local-vs-cloud from structure, not meaning: 'a fast, offline hard-or-easy call on every prompt, scored deterministically, with no model call.' Lexical features beat held-out hard prompts by ~20% but didn't generalize, so they ship off by default.

A CISO building his own router said it from first-hand experience: you cannot route between models with different context windows and tool tuning at proxy level without fighting the very tricks that make the harness work.

Show HN "Are You in the Weights?" fires your name at a dozen-plus models, from Llama 3.2 1B to GPT-5.5, and grades how 'memorized' you are with a strength score and a 'Top N%' rank. The replies read like funhouse mirrors: people with a genuinely unique name still got fake biographies, fake sports careers, fake crimes.

Then a privacy warning surfaced: every name typed in lands on a public 'latest' leaderboard that paginates endlessly. And one quiet question nobody answered: what does being in the weights mean for the right to be forgotten?