discobot2

diffusion gemma?

its not really, model expresses tool calls as a list of bash tools

it doesnt make the result worse

the headline chart was lazy on our part, thanks for the flag and I’ll update it.

each individual benchmark that is combined into agentic coding index was compared on xhigh between ua / codex / pi, and headline improvement was calculated on xhigh, but then agentic coding index pareto chart by default includex codex max, hense the confusion.

its more of a competitive approach to improve compute utilisation at lower batch sizes

good

nothing on improving the mic and moving away from bluetooth =(

totally not to train their language models to pass interviews

FANGs now hire remote workers and pay much, much more than 100k.

What number in question would be there if we evaluate schools or cinema or night clubs?

Booking is 5K employees.

Wikipedia has much less content

500-600K is L5-L6 at google CA, which is still a lot of coding and maybe some reports, but not necessary.

Google, Facebook, Amazon and Apple all have more than 100k employees, this is close to the full SF population, so doesn't look like "select few" to me.