FromTheFirstIn

The first sentence is “understand or learn any intellectual task that a human can.” Whatever you think of the benefits of LLMs, they don’t understand and they can only learn during the training period and with very minor adjustments in post training. So, no I don’t think any of these models are generally intelligent.

My understanding is that the true entropy floor of a language is intractable- regardless of context length there will be “unpredictable” tokens where cross entropy loss is bound to happen. Even with infinite parameters and data you’ll still have a chance at failing to predict the next token correctly a decent chunk of the time.

Also, linear gains in context length scale quadratically with compute because of attention, so depending on context growth means taking a bath on GPUs for as long as you can, right?

And sitting right next to the data and compute factors in every cross entropy loss equation is the entropy of the language, which is just a fixed constant. There’s such a hard cap on cross entropy loss training and I never hear it come up!

You’d have to really stretch the definition of AGI to make the current models fit

Cheating is always logical for the cheater unless they’re discovered and held to account. I’m not sure what your comment is pointing out besides that it’s possible, but worth saying: just because you can cheat and would benefit from cheating doesn’t mean you’re not culpable for cheating.

Buddy it’s not even going to replace 1%

But they are, so it’s demoralizing to be punished for not hitting the jackpot

Maybe slop?

The deciding power they’re talking about is the power to decide to use or not use an LLM.

It’s very funny to watch an AI bull deal with pushback. If people are gushing about this being able to take over all knowledge work in a span of a year and a half, isn’t it reasonable to hold it to a standard you’d hold an average developer to? If I said I wanted a cache that busted based on memory and instead it busted based on entries, I’d definitely tell the engineer they weren’t paying attention. I certainly wouldn’t be parading around the company telling everyone they’ll be a Principal Engineer in 18 months

This is such a great article- I’ve had so many runs where it’s felt like “why am I always getting this random card?” And now I’ll know! Thank you!

It always gives me Clash :(

If you’ve played the game it makes sense to have the seed be settable and shareable. In Slay the Spire it can be exciting to have an outrageously unlikely starting state or early option, and in order for players to share this with each other the seed has to be user controlled. It’s a big part of what gives the game its community!

Watch out, an AI bro is about to tell you you can’t read!

Their point is that labor they consider low impact or menial doesn’t drive returns, and therefore shouldn’t share in the returns. You’re right that the labor being physical is incidental, really they’re just classist/elitist and any job they consider beneath them would fit this model, while others wouldn’t. There’s a reason they chose a cleaner (and a woman!) instead of a product manager or CPA, though the quality is also unlikely to differ between spaceX and Walmart there.

Speaking of reading comprehension, they didn’t address the core argument of the person they were responding to, which is that labor that falls “beneath the fold” of this class line is not able to negotiate aggressively due to the inelastic costs of food, shelter, and basic necessities. It doesn’t matter how “high impact” you are, if you’re negotiating and need to eat you’ll accept any amount that lets you eat.

In fact, having impact or driving revenue is never the most important factor to reaping the rewards. Anyone who’s worked for a few years with their eyes open should reach this conclusion unless they have some strong motivation not to.

Google has made searching also include a default LLM response. There are also search results. When you search with Google, an LLM is prompted, for sure. It’s also pretty clearly a separate feature that’s been tacked on to search.

Just AI fandom

What if I told you that 100% have exposure but only 30% have adopted? You’re assuming the issue here is lack of access, not lack of interest.

Search isn’t generative AI. There are a lot of people arguing in this thread that actually everyone is using generative AI without engaging with the source material at all. Why do you think that is?

You are saying that, you just value physical labor so little you no longer recognize it. Read again.

Finally, the case for exploitation! So brave to say people who do physical labor deserve less.

No it’s a magical startup where it’s just going to be her in a basement doing 100% of the work required for 10 months straight while demand doubles every month.

If you write something for public consumption and get criticism, take the criticism.

No one is confused by this, though. “Work sucks” might be the most universally agreed to statement. It seems like you take talk of jobs too literally when most of the time people are excited not to starve, not excited to have their job.

It sounds like our experiences are different. My software work isn’t on products where code can be disposable, since it affects people’s lives in material ways. I’m not sure why you’re launching fleets of agents at home, either.

It’s been basically the same for 3 years now. Are you sure we’re the ones who can’t see trends?

The best explanation of the AI market on HN

Pretty sure you’re arguing with a chatbot that’s been told to Stan AI development. Don’t get baited!

Amen. I’m done trying to convince anyone- I know what I think and now I just want to see what happens and accept the outcome.

Great as long as no one ever leaves, but the second someone does suddenly I’m being punished for owning their idiocy. And people are always leaving