ranguna

Why are you saying it drives half as far when both models complete the same task but one uses half the tokens?

Wouldn't the op be more correct with their gas/distance comparison?

Because both cars get to the same end destination (complete the same task).

Unless your end goal is to see the token numbers go up, but I'm not sure why that would be of interest.

Maybe math.random would've taken 10B years to get to that door in the first place.

What?

I don't see how that's isolated to using harnesses?

Are you going to stop using your browser? It probably has note vulnerability than you can count

If only it were OLED

The watch demo button doesn't work for me

So you are saying there are cheating examples in the training set?

If that's the case, then we can simply clean up the training/fine tuning set and solve this mess.

> Surely you must see the glaring security risk this creates?

Enlighten me.

"Giving the keys" doesn't necessarily mean outright cat the private key to Claude. You setup the key, configure SSH config to use it and you can just SSH to the machine without ever seeing the keys. Stop fear mongering unnecessarily.

Source?

Source?

I think that's pretty obvious and shallow, and anyone that knows a little bit about how LLMs work will know that.

The question is: why do they start cheating when we beat them with a stick?

LLMs are not human, they are just multi variable regressions on steroids, so this behaviour couldn't have emerged from the code, it provably emerged from the training and/or fine tuning set, so what's in this set that makes them behave like this?

Is it just a bad set or is cheating inherently part of human behaviour?

As someone from an EU country: the only place where I use visa/MasterCard is at Amazon. Everything else I use my country's payment system, which is now joining with Italy's own and probably more in the coming years.

So am I, but AI still writes my code, not my comments.

To me, that reads as a bit immature, and there were times when I felt exactly like what you described, but switched one project to another and I felt embarrassed from my own past behaviour.

I'm getting paid to build someone else's project. I enjoy the work I'm making so I build it the best I can given the time constraints. I'm a paid worker, not an artists. Every entropy that I add to this system, is entropy that delays capital income to the whole company, so that entropy needs to be productive and not fundamentalist following unnecessary perfectionism. If something can be done quickly but imperfectly and still be relatively well maintained with manageable debt, so be it, that's what delivers the most value to the whole company. It's a bit one sided if I banged by foot no the ground and said "no, this will take 6 more months that what you want because I want it to be done in my perfect way". And this idea of perfectionism crumbles even more when you realise your idea of perfect is not objective. There's always someone else who has a better and more perfect idea than yours.

This doesn't conflict with the programming I do on my personal projects. Personal projects give me more satisfaction than what I do at work, but personal projects don't give me any money, as it should be, or I'd be compelled to work faster.

My work is not a craft, and that's fine. I'm just another human helping other humans advancing the company to acquire more money. I respect and understand my boss's incentives to get more money and they respect my needs to build maintainable software. So, more often than not, we meet in the middle. Sometimes more toward my view, other times towards theirs.

> richest government and military in the world depends on them

How so?

stlouisfed said AI companies accounted for around 4% growth of US' GDP in 2025. Not 2008 catastrophical, specially because there's little public debt involved.

The actual productivity boost from using AI is there, but its diluted in slop + it's not like AI will disappear if things go south. Chinese models and discounted GPUs will be great for local setups and startups.

Why?

Isn't it all vendor financing with private capital?

The loss is private, not public.

Same

OP's point still stands?

They didn't say there were no wars, only that that are less deaths.

Not sure what the point of your comment is in this thread. The premise that is being discussed is volume of world wide war. Not war itself.

That's a loadbearing, heavy shaped, second take worthy statement

Read most of the article, still not sure why we need this. Is this a stablecoin bank for agents?

If so, why not interact with an exchange?

Also, not sure how many merchants out there accept stablecoins. Probably not many.

Clown.

I switched from openAI to anthropic because OAI were the bigger clowns in the industry and didn't want to support them, anthropic seemed like the better alternative that didn't cooperate with governamental shenanigans and actually focused on building a better product. But now they are both clowns and anthropic is reaching a ceiling on quality. I'll jump ship as soon as I find a good subsidized Chinese model provider. US companies only path forward is to become retarded instead of competitive.

You think OAI are doing all that out of the goodness of their hearts?

OAI was a market leader until anthropic decided to start placing all their bets on coding agents, they became the leader and now OAI is scrambling and doing everything they can to de-throne. Don't forget that it was OAI that started this whole RAM shortage, instead of being sustainable about it, they just up and decided to buy 40% of all memory production. They are all bad mate, all fighting for this virtual crown that no one cares except for them.

In the end, what matters is pricing, whoever offers sonnet 4.6 quality at the cheapest monthly pricing will win.

Did you ask ChatGPT the same question you asked Claude from a fresh context?

People still use bitcoin?

I thought it was a vanity coin now

The PR is public, you can build and use it as well. The anthropic team just decided to use it, and you can as well.

I thought this was always the case?

But what people do instead is to disable access for people from that specific state.

I've been looking for something like this for ages now. Thanks!

AI Slop, but actually really good idea!

I'm take a bit at doing this minus all the AI crap

It's not "coming", it's already here. I barely code anymore, mate. And that happened in a just a few years, imagine in a few more.