But using the milking machines from A Boy and His Dog.

I used to be quite the AI skeptic, I even tried it once or twice but it turns out that most people refused to use AI and our collective wallet voting was killing it. Nobody seemed to care as it turns out, not even the AI employees.

It works both ways? ;-)

We already have specifications though, so that’s not different. What happens when the AI is wrong and wont let anyone deploy to production?

No one really. Code is for humans to read and for machines to compile and execute. Llms are enabling people to just write the code and not have anyone read it. It’s solving a problem that didn’t really exist (we already had code generators before llms).

It’s such an intoxicating copyright-abuse slot machine that a buddy who is building an ocaml+htmx tree editor told me “I always get stuck and end up going to the llm to generate code. Usually when I get to the html part.” I asked if he used a debugger before that, he said “that’s a good idea”.

I think LLMs didn't remove the need to verify code and shifted the failure modes from syntax mistakes to subtle design and security regressions that slip past casual eyeballs. In my experience with OCaml and htmx the right belt and suspenders are the OCaml type checker, ppx_expect snapshot tests for generated markup, QCheck property tests for invariants, and a CI step that runs html-validate plus a simple fuzz of input values to expose XSS or broken attributes. I've found the practical lesson is to treat LLM output like a junior dev, keep prompts small and testable, require human review for security or API design, and accept that you buy speed now but pay with maintenance debt at 2am.

I stopped worrying about that and just write code and do the debugging. Why pay for speed when the cruising lane is perfectly acceptable?

This is something I've been wondering about...

If boilerplate was such a big issue, we should have worked on improving code generation. In fact, many tools and frameworks exist that did this already:

- rails has fantastic code generation for CRUD use cases

- intelliJ IDEs have been able to do many types of refactors and class generation that included some of the boilerplate

I haven't reached a conclusion on this train of thought yet, though.

Pre-llm corpos my thoughts were that we should be training juniors on code generators. Instead we’re somewhere between rtfm or dont.

Because it was always a vapid distraction from life.

I see a future just like the seo issue of today, where the well is poisoned and llm information is garbage.

Now why would anyone publish for free? Publish or die is dead.

You can shape the AI responses for some niche topics relatively easily even on accident. I recently saw 2 people arguing on a forum on a very niche industry topic and one of them started to use Gemini as a source to argue and Gemini was already referencing their thread as a source. I'm imagining people could start doing that on purpose with their own astroturfed blogs and public social media accounts.

> one of them started to use Gemini as a source

We need a variant of Godwins law to reflect (and prevent) the use of AI being used in internet squabbles.

Cut communication and notify them you dont like speaking with their llm and expected the conversation to be between the two people. Discount their credibility.

Apple will utilize them when needed and scoop up extra capacity after the bubble burst.

Did you read the article? Apple's servers are M2 Ultra class and not able to run modern models.

What does that matter? They cant be reused for other things? Nonsense.

What other services does Apple have that people would be paying for? The ones they have today are either iCloud storage, which does not need much compute, or merely an alibi so they can claim with an almost straight face that Apple's "Services" revenue isn't basically just the App Store 30% tax. That also explains why they are constantly shoving ads for News or Fitness our throats in the Settings app.

And no other services will exist or grow?