You made it sound like any country that wants nuclear weapons can get them.. just look at Iran. I think existing nuclear weapons club is very interested in keeping membership pretty tight
Same experience on M4 Max .. but quality of qwen still leaves so much to be desired after getting used to virtually unlimited tokens at work. Many people on this (and similar) thread seem to believe local models would inevitably improve, and I want to believe this too, but I don’t see this ever happening without growing in size
The legalese around it is that it’s always contingent on your employment on those futures dates. It’s based on past performance, yes, but it’s not payment for past work (that’s what bonus is for), its incentive to stay longer and contribute to company’s success. When you are out of the door (for whatever reason) company looses need to incentivize you to:(
It’s not a downside perhaps but a tradeoff. Implementing internal sampling profiler is not impossible in general (though it will depend on exact python implementation) but it would impose different overhead on the target and might skew results a lot more than external one.
Even when performance is not critical it’s possible to write (or vibe?) disastrously slow code, so having profiler handy is always a plus. It might be a deciding factor between “we must rewrite it all in Rust” and “oh we added exponential complexity by accident” and save a ton of time
Given that EU tech salaries are a lot more tame, it would be interesting to see how 600m were even used. Hopefully there’s some good R&D there and not some French alps retreats and Porches for founders