P.S. O1 is really just my excuse, we are working on scaling laws stuff and it is something we keep thinking about The STLM project is amazing @LeonGuertler et al. (need handles...)
@BlancheMinerva you wrote about it but I can't find it (I remember the bottom line of we are not even close, but I want the references evidence estimations and deeper thoughts) @yanaiela @ShayneRedford @natolambert thoughts?