saberience

Ah yeah, Carbon it's so OBVIOUSLY a superior element to Argon!

Mythology implies gods etc, Fable implies a something more like a fairy story.

That's fairly obvious to me. Most people associate Myth with the gods, which are kind of the biggest and most powerful things we can imagine.

Not so much of a stretch...

The Google naming convention is so nonsensical it's comical.

With Anthropic and OpenAI, it's obvious immediately which are the better models. I.e. Sol is better than Luna, or Mythos is better than Fable.

But Google, let's pick Argon, which to most folks means mostly nothing. And then... Carbon! So which is better, Argon or Carbon? Nobody fucking knows. Plus neither sounds cool or viral or mysterious or even provokes curiosity.

When Fable was first leaked it sounded different and intriguing... Carbon just sounds... boring.

Except the results are AI slop that no one actually understands and make absolutely zero difference to the world. I.e. research with zero practical applications.

Yay, the slop generator made 2000 pages of Lean which no one understands... to solve a problem with no practical results for human-kind.

Great, well done OpenAI.

OpenAI releases some new PR slop which actually makes no difference to anyone's lives, except the Mathematicians, where it ruins their jobs.

Nothing.

How are they a mess? They look amazing honestly, and I work in an office surrounded by experts in analyzing statements like this.

I don't trust these Artificial Analysis indexes and I'm not sure why anyone considers them authoritative, the site just sprung up a few months ago and now everyone quotes it?

From my own use of the models I find their ranking to be totally inaccurate.

Oh no, the poor sheep!

Why are sheep so important to protect versus wolves? They are both animals that existed in nature and can continue to exist in nature.

Why should we be morally inclined to try and "save" sheep at the expense of wolves. Sheep and goats and deer survived throughout history without the help of humans they will continue to exist, with or without the help of humans against wolves.

I stopped looking at Pelican pictures about a year ago when I realized they mean nothing.

Yeah but if 99.9% of people would get the bike totally wrong, why does it matter if the LLM doesn't get it perfect.

You're nit-picking on a detail no one actually cares about.

Isn't this a good thing? I mean, is there a direct correlation between electricity usage and country = better?

There's a difference to pushing out models and having people and more importantly, enterprises, using them.

There's a ton of good models out there, but most are not making any money or profit.

So let's see what happens next. But I don't see any or many enterprises switching from using their Anthropic or OpenAI models to use Meta.

I see this from everyone I know who works at Anthropic, they seem trapped in some sort of AI psychosis where AI is on the verge of ending the world.

Meanwhile, back in the real-world, most people I know use Chatgpt a few times a week for totally mundane shit and even more people never use AI at all.

And in my world, the latest frontier models still do moronic shit in the codebase all the time.

Meanwhile, my friend X at Anthropic tells me he fears that AI is going to have a terminator moment within the next year or two...

It's bizarre honestly.

Also, Opus 5.5 is considered Anthropics best model right now and if you try it, is better than Fable in almost every task.

You're also splittin hairs if you argue Opus 5.5 isnt "frontier".

https://artificialanalysis.ai/models

Artificial Analysis considers Opus 5.5 to be the best current model, yes better than Fable.

You realize that the models are not deterministic right? It can literal vary from attempt to attempt.

Yes but they're not under-estimating how much worse it is than Opus 5.5 or Astra...

Amusingly enough, I corrected it, and it doubled down on there being 29 rs until I forcefully told it that it was wrong... :)

I just got Claude Opus to fail at counting letters in a string.

Me: What about in rrorooroororororrrrrrorrororororororrrorooorrorroror

Opus 5.5: There are 29 r's in that string.

There are 30rs in the string. If you want I have a screenshot.

The reality is this, the models "think" in tokens, not in letters, so they can easily make mistakes like this and it's very hard to stop these kind of dumb mistakes happening. So yes, frontier intelligence at work.

Ed Zitron is a hack with no tech or financial literacy, who gains views from fellow biased AI haters.

His "facts" are not facts, he makes stuff up, miss-states things constantly, and is financially illiterate. Whenever I read anything he writes I'm constantly face-palming because it's basically all rubbish.

He's anti-tech, anti-AI, incredibly biased, and just honestly full of bile.

And yes, I'm saying his data is wrong, his interpretation of the data is wrong, and his conclusions are wrong.

"If OpenAI had a sound business plan and was on strong financial footing, they'd have IPO'd."

This is a fallacious assertion, there are so many big private companies with strong businesses that stay private for a whole variety of reasons. E.g. Stripe, Revolut, Databricks, etc.

You're letting your bias leak out here.

This isn't using any frontier pure math graph theory results, or anything like the kinds of papers that OpenAI is producing.

The math in the linked article is understandable by any undergrad in Maths. Hell, I can follow the maths in that paper, so it's nowhere near like the sort of abstruse stuff in the OpenAI maths papers.

And again, if you look at some of the problems in those papers, it's clear they have no practical use to make our world better.

I.e. proving that this problem which we thought would make 1 million years to solve, actually has an upper bound of 999'999 years of time. Ok, cool, very interesting, but actually has zero practical impact on our world.

Because they published a ton of slop papers with terrible English, impossible for humans (even experts) to understand, full of non-standard terms, invented jargon etc.

So effectively, stuff got proved, but people don't really understand how, so it's mostly fucking useless and done for OpenAI's marketing team, while also pissing off the maths world at large.

None of these graph theory results lead to any applications in the real world.

But sure, let me know when they do. I'll be waiting.

I'm not sure how you define "knowing how the world works", but knowing that a very very niche algorithm upper bounds that we thought was x^100 and now we now it's x^99, isn't that interesting. It doesn't really tell us much more about the world and it doesn't have any applications for our day to day lives.

It's interesting to theoretical mathematicians only, for anyone else it's just noise which doesn't affect our practical day to day reality at all.

In fact, every one of the results is basically just novelty crap as far as the world goes.

Let me know when AI discovers the cure to cancer or aging etc.

They still have issues with problems like this actually, and I use all the frontier models from all the major labs, so it's not solved.

I agree with your premise though, we require specialized humans to generate the highest quality data to improve the models, and it would be more impressive if the models could do this themselves...

Anthropic have done experiments with having "dumber" models try and make a smarter model, smarter, but it still seems tricky so far.

But if you think about humans, we have the same issue. If you have a super smart person, how do they get smarter? Can you have dumber people make a smarter person even smarter? It's a hard question.

Who made the formal proofs and who checked them?

I mean, Mistral is about 9-12 months behind here when you look at its overall benchmarks versus the models released around a year ago.