Hacker News Reader

Facebook released a pretty impressive SOTA system with public data sources a few weeks ago

Are you talking about wav2letter@anywhere?

If so, I haven't dug deeply yet, but I suspect the author's point stands - inference-time efficiency is one thing, but training compute (and separately, data-efficiency) is another.

I don't think the author is necessarily criticizing anyone, and I don't think he/she is saying that these tools/models/papers/etc aren't welcome. My read is that "pushing STT beyond the FAANG use-case means we should stop confining ourselves to FAANG-scale data and models".

Particularly https://github.com/facebookresearch/wav2letter/blob/master/r...

It uses a publicly available huge training set

What does "SOTA" mean in this context; all I could find is "Software Over the Air" but I'm unsure of what it means in a Speech to Text context

“State of the art”.

But I think your original question is still great:

What does state of the art mean in this context?

It means "State of the art", which would be roughly equivalent to "as good a result as its currently possible"

Reply on news.ycombinator.com