geraneum

It largely works and it's a massive business success.

The engineer is suggesting that it could be done cheaper and maybe with better outcome. Ironically, this is a classic business case.

If you’re puzzled as to why this exists, imagine that, out of the goodness of your heart, you donate $230 to OpenAI to support their mission of rear ending the singularity, and receive Codex Micro memorabilia as a token of appreciation.

It’s because they don’t pitch to the users. They pitch to the investors. Have noticed how everyone and their dog now has an “AI story” on their website? Yeah, because they won’t get funding otherwise. And these are not even the AI companies!

I have this funny feeling that someone’s probably gonna ask their favorite frontier LLM about the fruit salad thing to refute your point.

Yup, my mistake.

deep understanding is going to become a commodity soon

How does the text generated by LLM make “our” understanding deeper compared to text written in the books?

Claude has recognized some important semantics that we ourselves lack a good word for or at least under-appreciate.

Ah, I love when Claude reads our collective minds and fills in the gaps to address the load-bearing seams genuinely with an honest caveat.

I don't need a vendor's special-sauce computer to build for consoles.

This is ironic. Of course you do. You're already using it. You just happen to like the Microsoft's walled garden and not Apples. That's fine, but let's not pretend that "Windows centric" is anything but another monopolistic and exclusionary ecosystem. If you want to use Nvidia drivers, or develop against DirectX, can you do it on a macOS? I get the criticism of Apple's walled garden and I'm against it too, but not from your position.

I don't disagree with you, but we're talking about affordability which is a different topic.

It's especially nice when you work on projects that aren't iOS/Apple native and you don't get Mac workstations.

If you develop natively for macOS, then you probably need the OS for it. It's the same for Linux and Windows. But building is not the biggest hurdle because you can build in your usage-based-billed/free CI as well the same way you do for other platforms.

For complete native development (system APIs, OS interactions, native UI) against macOS/iOS on PC, Apple needs to open those OSes for PCs/Pixels/Samsungs/etc. This is a software issue, not hardware. Otherwise Apple silicon Macs are less restrictive than PC hardware in general!

Never fake success. Run builds, tests, linters, and relevant checks whenever possible.

Don’t make mistakes. Don’t lie. Be successful. Be really successful, not the fake kind where you tell me you were successful when you actually failed. Know when you failed. Don’t fail.

People not having access to tech is bothersome, but nowadays Apple offers affordable machines compared to the same/worse specs from competitors. For example a MacBook Air with Apple Silicon and a moderate amount of memory.

What I have been missing in all this debate is substance.

That’s strange because in this (and Andrew’s in lesser extent) post there’s plenty of substance on both technical, management and corporate influences like the difference of styles guides vs agent instructions, binary size, compile time, Anthropic’s marketing and incentives, etc. It’s hard to miss.

Do also frown upon people working in nursing homes? Or your uber driver? No one frowns upon the people who work/help people who need it.

Hire not enslave! Like how you might hire a driver, or housemaid, or anyone to do a salaried or contract job? How does that perpetuates the problem?

If they pay her what she’s owed and the damages. She can get her place, hire people and pay them to care of her or help her.

Maybe in previous failed attempts that what the model landed on and they’re preemptively stopping it. Did they release the any info on the failed attempts?

The article reads like a compilation of online chatter, social media post, etc. on AI. Not surprising to find unsubstantiated imaginary numbers there.

In the farm?

It's just standard nerd drama.

The elephant in the room are the trillion $ companies having a horse in this race.

That sounds like a good idea but how do you know what you’re shipping?

It's so sad that Rust vs Zig has been dragged into the AI psychosis vs anti AI narratives. I feel people are taking sides or picking and choosing now based on their allegiance and religious values. What good comes out of this?

non-deterministic problems

I'm guessing you mean probabilistic? Nevertheless, you have an indeterministic variable here which is what the LLM generates.

You have to put similar amount of resources when writing in Rust as well. With the difference that it’s more front loaded. Personally I’m a fan of Rust’s approach but the price for having bug free code has to be paid, regardless, one way or the other.

Not really. It seems like the Rust rewrite blog post triggered this. Which, even though I don’t agree with the tone, seems to be valid criticism.

and the project leaving isn't as good for Zig as Andrew is making it seem

In this case, Bun was acquired by Anthropic. Leaving Zig is not necessarily out of merit. As much as they pretend it was.

gpt3->4o->sonnet4->opus4.6->fable

While there are a lot of resources being poured into improving and integrating tooling and use cases and increasing model sizes, the leap between the models you notes are not as much as you think it is. It's the same tech that's scalable to a certain limit of diminishing return. It's a recurrent pattern in technology.

yes, I assume we will have the 'replicator' from Star Trek, and programs will just appear as you want them.

Yes, perhaps one day. Probably no other "job" as they are today will have the same shape and form. Cancer will be cured and fission will be solved. We may even crack teleporting. Let's hope these all happen and the outcome is not detrimental the livelihood of humans.

Cue some story here on a bank or airline somewhere still relying on cobol backend servers.

There's existing money and expertise in those environments to rewrite the whole thing, yet they don't. You may loan them free engineers/experts and they might still not rewrite anything.

Well I was agreeing with you. Skimmed the paper as well.

A subset of python is python. Half a tomato is still tomato

The funny thing about this is not that the first sentence is wrong, which it is. It’s the failed reductio ad absurdum.

I’d prefer a more direct claim in a paper

This is not written to be just a paper. The target audience include media and online forums, and then maybe academia.

Edit: typo