Most LLM calls in production aren't really generation. They're decisions: route this ticket, flag this prompt, score this urgency.
Laya handles those in one forward pass: ~33ms, calibrated probabilities, zero output tokens.
Open weights, Apache 2.0.
https://t.co/yJmHwyE5Un
Big news: I'm taking part the Vinnies CEO Sleepout to raise funds for people experiencing homelessness. Support me today to help thousands of Australians in need. https://t.co/bxyBJd6Egj
@martinfrogers One thing that will change after this is the cash economy. All the government support is going to people who have been putting income through the tax system.