Nimitz14

Great question. I think these people are good at spinning narratives _and_ the people above are not doing the work they should be doing to check whether the narrative checks out: I think good leaders (and every manager below) should be doing inverse skip and skip-skip meetings often enough to have a good sense of whether what their reports are saying is accurate. But most dont

You have some fundamental misunderstandings on how LLMs work.

You’re not getting it. The probabilities do not change at all. The only change is given some probabilities there is a deterministic method for determining which symbol was sampled from that distribution. The distribution or sampling process itself is not modified.

Uh isn’t google known to have the best tooling in the world

Ai slop

He said it nudges it to be more deterministic. Your comment is not correct.

This is mostly slop. But you may be directionally correct

Really really cool. If they can serve this efficiently it would disrupt a lot of things.

I think your own math leads to the conclusion the public apis are not serving models of that size. They couldn’t afford to

3/4 are actually negative value for most customers

Your answer presumes we cannot hold people accountable. I think that is incorrect.

Most of these errors will not be meaningful. Real speech is full of ambiguities. 3% is low

I think you're bang on the money fwiw. But also worth mentioning that it's OK to ask rather than trying to predict and feeling that having to ask means failure

This is clearly a partisan website (see its name). Read with a large grain of salt.

Yann was in charge of FAIR which has nothing to do with llama4 or the product focussed AI orgs. In general your comment is filled with misrepresentations. Sad.

Sora is about data gathering. This take is too simplistic.

There's still too much fat, or rather people who just add net negative value. My org would benefit from firing easily 50%.

I'm surprised at the model dim being 2.8k with an output size of 200k. My gut feeling had told me you don't want too large of a gap between the two, seems I was wrong.

This is normal. Rope was introduced after bert/gpt2

What is it like living a life completely ignorant of reality?

If you'd like to learn, feel free to ask chatgpt on the leftish pushback against abundance. Or historical examples of leftists blocking housing projects. Or environmentalists prioritizing niche interests over those of the general community. There are many, many examples.

We're all gonna be PMs.

Think the microsoft terminal saga shows a pretty clear track record.

This is very good. I've also noticed I do the polite engagement sometimes and how its roboticsness is obvious to any person with not low EQ. Some people seem to just be naturally happy in almost any settings and its a very valuable trait imo.

[flagged]

So this is not the case for https://gemini.google.com ?

If your friends are the sort that stick to tribes instead of thinking independently get smarter friends.

Not the same thing at all. Corporations care about their data a lot and would cancel deals over this. Noone cares if some authors get upset, they have no leverage. Disappointing how people will make confident statements while being so clearly clueless.

Any continual learning papers you're a fan of?

If you had spent the time writing this and other comments googling/chatgpting you could have learned why returning to the gold standard is a joke.