istewart

I feel like it’s too much to hope that even the HN crowd have figured out Balaji S. is a more prolific bullshit machine than any LLM…

Wait for AI and Crypto 2.0 to burn out, we’ll get there

oh no, son

you built a talking cartoon butt instead

we’re ruined, son

I suspect it’s a bit of a tell that upcoming hype cycles will be focused on biotech. Not that any of these people writing checks have any more of a clue about biotech than they do about computers.

My baseline understanding is that “NPUs,” as such, are vector accelerators with perhaps lower precision and definitely lower peak TDP. I say this because much of the incremental ML research I’ve skimmed over seems to be around getting away with lower precision, dropping down to FP8 or even FP4 from FP16 when they can get away with it.

I’m still confused as to why and how this is an acceptable tradeoff to firing up an iGPU with precise power/TDP stepping. Perhaps one of those situations where the power budget and latency to fire up the whole GPU block or burst it to max power ends up costing as much as the actual calculation. I think for purposes of this discussion, we also need a source that sheds light on the architectural differences between NPUs and GPU shader/execution units.

Deep dive into complex chemistry may be the most important thing in this thread. The bullshitters need more pushback like this, even though the effort involved means it can’t happen nearly as often as the bullshit. Thank you.

I’m guessing these people were JAQing off hard enough that they got kicked out of their local Oxford-style debate org and had to start their own.

“Nerniacular framen, dude!” - Bart Simpson

Has this person turned up shilling their book on Coast to Coast AM with George Noory yet? If not, I think it’s a lock for 2025

She’s still a go-to for celebrities who want to be seen as telling raw, difficult truths; Harry and Meghan obviously being the flagship example. They’re surely trying to piggyback off that here.