mobyduck648

It depends, if Anthropic keep nerfing Claude’s capabilities they’re going to lose to their competitors. I’ve been doing defensive cybersecurity work for the last couple of weeks and Claude, a tool which should be ideal for automating the tedious parts of this, has been a complete pain in the arse.

Claude: I’ve found a security bug in your codebase that nobody knows about!

Me: Oh shit that’s not good, can I see it?

Claude: <blocked by safety classifier>

Repeat ad nauseam.

The prissy little bastard’s recalcitrance made me to do a lot of stuff by hand anyway, but instead of being bored I’m now actively irritated by the bot telling me what I can and can’t do. Obviously can’t use them at work for regulatory reasons, but when I was doing similar stuff at home to secure a side project which has to live on the public internet GLM and DeepSeek have no such qualms and honestly I’m convinced the whole ‘DeepSeek is the budget option’ thing is mostly behavioural economics; the cheapness to me feels like efficiency gains more than nerfing the model.

I wonder if AI will leave much progress behind when it implodes.

I think it will, it’s the one bit of AI optimism I actually have. I have a theory the big AI firms love for example Yudkowsky’s doomsday cult so much because it’s easier for them to imagine the end of humanity than it is to imagine them not getting an extractive oligopoly over an emerging technology.

The tech itself is actually decent as tech, it’s not what the boosters claim it is but even if development stopped advancing today there’s loads of niches it’d be changing for a long time; computer vision in medical settings for example. Useful things like TTS will be much better going forward than they ever were, and for better and worse things like computer vision in general (ie in public) will never be the same. LLMs are basically magic NLP boxes too, they’re useful for classification as much as generating slop.

With that motherfucker it’ll be something vile or the kind of concentrated cringe that makes your earwax melt just sharing a biosphere with it.

Probably both.

They need a good humbling, and their own hubris will be the thing to do it.

I’m keen on local models but prices are getting out of hand. I insured my workstation recently for more than twice what I built it for, because it’d cost that to rebuild it with identical parts.

If an AI was provably 100% correct then I guarantee you people would listen to it.

Intelligent people will believe the most obvious nonsense sometimes; engineers are represented among young-Earth creationists for example.

To be fair, it’s going to be more like running an electric kettle constantly which isn’t going to excessively load the grid.

Unless you’re in the UK of course where the grid operators have to manage capacity around half time of major football games as everyone puts a brew on at once.

And now the oligarchs behind the big AI firms ingest pirated scientific papers at enormous scale to build their products, and they see fuck all consequences for it. It’s quite plausible the very papers Swartz was hounded to his grave over ended up in the training data. Justice is a fucking rare thing.

Don’t fuck Spez actually, there’s no doubt in my mind he’s a terrible shag.

I’ve come around to this poster actually, next time people moan about them I’ll start using the long s!