alentred

When you change the language you are essentially moving to a different Wikipedia site, so yes it does - it pulls the results from the local site you are on.

Same for "AI". LLM is about the only kind of AI we talk about now, so the two blended. Nothing dramatic I guess, but it bothers me when tech people, who are supposed to know the difference, confuse the two.

> My office was wired with two sets of electrical sockets, one connected to grid mains and the other connected to the UPS (for important equipment only).

This I think is a standard practice in any large office building in France (Europe? Global?) too, regardless of the frequency of the power cuts. I don't remember witnessing any power cuts in the office (not since long ago anyway), but in most of the places there are systematically white and red sockets, red ones being powered through a UPS. So, server machines go into red ones. Well, we don't have servers anymore, but the sockets are still there anyway. Hopefully UPS too, LOL.

Does anyone have a real-world usage feedback on this model? Because the benchmarks look, I would say, too good: it scores even higher than Qwen 3.8 27B in some benchmarks - I am not even sure how is it possible. So the question is if this is benchmaxxing or is it actually good for practical usage like coding in complex projects, research, etc.

I will write you every day.

(I suppose people downvoting this don’t know the reference. Way to make us feel old :) )

Thank you, sir.

I am still not sure if the joke was bad or if everyone takes everything too seriously nowadays.

Don't give 'em more ideas! :)

Oh boy, this is sooooo addictive.

Look, I don’t mind it but let us know - are you training some model on our inputs ? :)