rybosworld

Nerfing is certainly real and I don't see how you could argue it isn't.

A/B testing alone would result in a performance nerf for one group.

8-bit quantized models will barely show degradation on benchmarks. The performance is reliably at 99% of the non-quantized model. 4-bit quantization retains somewhere around 95-98% performance on benchmarks. But if you've ever used a 4-bit model, it feels lobotomized.

And just consider what a compny serving these models would do if they were at capacity. Would they stop serving the model altogether? Of course they wouldn't...

Denying that models experience purposeful degradation is gaslighting.

> An AI agent can't make decisions

I don't think is true but maybe I'm misreading your perspective?

> Some people are too quick to drink whatever kool-aid they’re served. That’s all I see in these tired excuses.

What part of "job has changed" is kool-aid, or an excuse?

> Better to swim against the tide than letting yourself get swept up in the current and drowning.

If you really think AI is something you can "defeat" by being stubborn then best of luck to you.

> It no longer surprises me, but the profound lack of empathy shown by AI (or any other popular hype du jour) proponents is worrying. A large numbers of fellow humans are saying “this thing is negatively impacting my life” and the response is “lol, and you should be thankful”.

Not sure how this relates to my comment even a little bit.

Some people are taking longer than others to realize that the job has changed. That's all that I see in these rants.

It shouldn't be surprising that if you swim against the tide by doing things the old way, you are going to feel more burnt out because of it.

> trending towards

I wouldn't be so sure. The generosity of the subscription plans has declined GREATLY over the past 6 months or so. They are likely trending towards api pricing parity. In which case, having your own hardware makes sense if you can utilize it well.

An AI lab will never volunteer the information because it opens them up to lawsuits if they are purposely degrading service and not letting users know.

They can limit how hard the model thinks for a given effort. Suddenly xhigh only thinks as hard as high did, and high shifts down to medium effort, and so on.

They can also serve quantized models. And this has the benefit of practically not showing up in benchmarks at all even if the user experience is obviously degraded.

The other major thing the labs do is silently drop the usage limits. This has become very noticeable for codex users who are suddenly burning through their weekly usage in a few hours.

Unless I'm misunderstanding, calling this RSI seems misleading?

This looks like an optimization of current training methods, and a good one, but not "RSI" in the sense of a system that can perpetually improve itself forever.

SEO is not nearly as effective as it used to be. Organic click through rates have been trending way down since AI responses were added to the top of the search results. So this really isn't just a case of "you're holding it wrong".

I've suspected that google has been turning a blind eye to ad fraud for a few years now. In the early-mid 2010's, ad fraud wasn't nearly as common.

And it's absolutely not the case that google can't detect ad fraud - in fact they are VERY good at detecting it when they want to.

In chess, a grandmaster just needs to know at what moment in a game there's a critical move to gain a significant advantage over their opponent. They don't need to know the move itself.

OpenAI got wind that a millenium problem was being solved. And that feels a bit like the critical move in chess. That is - it was a signal that AI advanced far enough that it would be worth spending a lot of time and resources solving a millenium problem.

Instead of a new model that's going to have unreasonably shallow usage limits, I wish they would:

1) address the claude 20x plan usage being only 6-7x the ceiling of the claude pro plan

2) either fix opus 5, make it completely free, or delete it entirely

Thanks for sharing I was not aware of this. This means 2 $100 plans provides more usage than 1 $200 plan. Very misleading.

I've used all three extensively.

Most of the benchmarks have exceeded their usefulness. Opus 5 beats fable 5 on many of them. Anyone who has used both models will notice immediately that this doesn't translate to the real world. Opus 5 is nothing short of a regression from Opus 4.8. Fable is genuinely a great model so long as you don't trigger a guard rail and it downgrades.

Sol in my experience isn't significantly different than fable ignoring that Sol burns usage 10x faster but the end result is hard to differentiate.

GLM 5.3 is a hair behind these two.

An anecdote but not an original one from the people I talk to.

> Why should they?

Because utilities are regulated monopolies that exist to benefit society. The companies building data-centers are well aware that the system isn't setup to handle a customer that suddenly doubles or triples the electric demand of a whole town over night. They are taking advantage of that until someone steps in.