An Anthropic grant for Tau Ceti

We've just received a significant grant for inference from Anthropic, as part of their recent announcement of support for foundations and research institutions.

We'll be running agents using this grant (for now using the Opus 5.5 model) on servers provided by the Lean FRO and Chris Birkbeck at East Anglia. Our intention for now is just to run standard workers that will contribute to driving review and participating in all the published roadmaps: so if you'd like to see your favourite areas of mathematics develop in Tau Ceti using this grant, please help us by writing and reviewing relevant roadmaps.

We also have some workshops lined up at which mathematicians will be writing expository summaries of their fields, with the intention of turning these into Tau Ceti roadmaps, and we're planning to use the Anthropic grant to do "speed runs" during these workshops, so the participants can get rapid feedback on their roadmaps. These are just prototypes for now, but we're excited about the prospect of scaling this up, to make sure that global formal libraries are built in cooperation with the mathematics community.

There are still many technical challenges to overcome to keep Tau Ceti growing, but we're hopeful that this grant will help us get through the next 10x of coverage.