quikoa

Overall I do believe it is a productivity win for software. However, the insane amount of money spend on AI (training models, building datacenters and even usage in some cases) is ridiculous. Hopefully that will get lower soon.

Datacenters contain few jobs so unlike manufacturing this doesn't look like a loss.

It's not productive because that productivity could've gone into something else if this wasn't necessary.

Great, that means fewer bloated JVM based products will be released as well.

> Apple does let apps download JavaScript updates like this, which is how tools like Expo Updates work, as long as the update doesn’t change what the app is. Whether changing your pieces based on ad prices fits that is up to Apple.

But there is no way for Apple to verify "as long as the update doesn’t change what the app is" right?

> Also, single bits are rarely used for booleans because it requires more CPU power to extract a single bit.

Not necessarily, on modern CPUs memory access is often the bottleneck. If you have many bits storing these in a bitarray can be quite beneficial for performance.

That's an additional cost to the user rather than infusing more resources from the provider like a subsidy.

This is why they enforce the use of client apps like Claude Code/Codex (although OpenAI is a little more lenient). Trying to create a network effect.

Good that they put rogue in quotation marks because there is just no way that this is some sort of accident.

Wouldn't it be trivial to detect benchmarking if the same requests are running on a fixed interval?

> Yes, well, so can humans given the "right" circumstances, e.g. "enterprise" software

The difference is scale. Letting LLMs go out of hand is like the movie Fantasia. Sure people can make a mess on their own but with LLMs can do so much worse in such a short amount of time.

> Generative models are now doing the same to coders

Programming involves quite a bit more than just typing. However, typing the code does help with understanding. Once we stop understanding code because it's just generated AI slop good luck fixing the mess.

Now that even Anthropic and OpenAI called for slowing down. It's obvious that we're past peak hype and realization of the cost of LLMs will set in.

LLMs can produce some incredible bloated and convoluted pieces of code if one isn't careful. Even to the point LLMs become useless. Blindly using these 'powertools' is a recipe for disaster.

We'll already see the hype to start cooling down. If you're okay with cleaning up messes you need not be so pessimistic.

"Others do it too" is not really a counterargument when Anthropic should've stopped this. They are yelling "AI is dangerous" the loudest so the least they can do is get their own monitoring in order.

Are you sure you don't mean revenue? Given the enormous investments profits seem very unlikely.

> What is more difficult, building a military missile or the software that runs it?

"If software was easier compared to building the a missile then Anthropic is in the clear or something." Is that the line of reasoning to protect Anthropic from any wrongdoing?

Quite a bit harder or even impossible to prevent software being used illegally that runs on clients than server side services no?

> easily outcompeting any company

If it were easy they'd be doing that instead.

If Anthropic/OpenAI would've shown incredible results using these models then I'd be worried about this. Instead we get Codex and Claude Code, bloated and disappointing software. I'm sorry but "use them in-house, and eventually eat the whole economy." doesn't appear to be a real concern with these two companies.

Yes and winning from Google in small claims court is definitely possible: https://www.huffpost.com/entry/why-i-sued-google-and-won_b_1...

Edit:

AI can be a great help here, ask it:

- How to prepare your case for your country.

- Where to file and how much it going to cost.

- Help you find weaknesses in your case and what you can do to fix those.

- Rewrite your text in proper legalese (I do recommend to prepare the your arguments yourself to avoid over generalization from an LLM).

Of course you need to verify because of hallucinations but showing up is half the battle here. Big (tech) corps + internal bureaucracy makes it quite expensive for fight these cases. Especially if there are many different countries.

There is no need for a copy. If another can do most of what OpenAI/Anthropic can do at a fraction of the cost (like the Chinese models) then the moat evaporates.

There is also a predictable relation between the input and output of a compiler w.r.t. the semantics of a programming language. Natural languages are ambiguous leaving room for the implementation to diverge that may not be obvious at first glance.

There could be a large overlap between OpenAI glazers and Star Citizen dreamers.

Exactly, they scrape the internet without any regard for copyright and now someone is surprised it happens to them.

What's next, subscribers believe they are paying customers instead of sponsored data providers?

So why doesn't Anthropic write a good Claude Code in Rust already? Why don't we get a Codex with the polish level of Sublime text editor?

These companies have many billion dollar investments and yet their main product is a bloated meh.

On the off chance that OpenAI executives are reading this. I'll totally believe AGI is here if they do this.

That won't even require an AGI.

> (at the moment)

That's why. That future is uncertain. So why gamble your future on something that's popular at the moment for something that could change completely a year later?

A language without a large stdlib pushes this that functionality into (transitive) dependencies. I hope more language will adopt batteries included approach.

It'd definitely need a solid team behind this and not just Evan Czaplicki if I were to trust a database with my data.

> something dangerous like AI model developement

It's very dangerous to Anthropic's bottomline!