When I read 2.5 miles I thought that's... not very long? I'm sure I've driven through much longer tunnels in Norway. Turns out I have: https://en.wikipedia.org/wiki/L%C3%A6rdal_Tunnel I had no idea it's the longest road tunnel in the world though, that's pretty baller.

We (Synqly) have a production MCP server which several of our customers use in their production services. The functionality ranges from mundane management capabilities such as walking through how to add an integration in natural language (e.g. support bot) to more than exciting use cases such as making complex queries across a number of tools to create aggregate reports.

It's a very popular part of our platform.

Yes. Two wrongs do not make a right.

If the results of the distillation are released freely for everyone to download, I wouldn't even call it a wrong.

Big models are built by scraping the recorded thoughts of everyone, so giving everyone a chance to run a distilled small model is just going full circle.

Obviously the underlying motivations aren't 100% altruistic, but I'll still take it.

I find it to be hit-and-miss, but overall more useful than not. What I like most is its relative brevity, and the response often helps me find a better prompt to ask so that's good. It's also really fast which I appreciate.

I guess I should mention that I'm using the deep research assistant, not the quick one.

We're currently in the process of restructuring our documentation and I'd be very interested in an early draft of this, fully understanding that caveat emptor. Is it published anywhere currently?