[hollowturtle]

Please do not cite Dunning–Kruger effect at random.

Who needs to generate a dumb demo of a 97% done crud app? We had code generators for those, everytime I read claims like that and I ask to explain further I then discover it's people who were not productive before generating the so called "MVP level things to completion with ease".

If you're trying to solve a HARD problem people REALLY have, it's a novelty that agents can't help with, otherwise if it gets 97% there MAYBE it's just a signal that your idea isn't that novel!

LLMs can effectively validate your business idea

[AussieWog93]

I don't really see your point. Most problems that people have aren't really super-novel, but just extremely bespoke.

To give a specific example, 12 months ago I had a client pay me me to make a Chrome plugin that changed the rows in his Shopify Products page to display Quantity and SKU.

These days you'd just one-shot it in Claude.

[hollowturtle]

First of all it just underlines how shitty the web has become, second If that's your work I'd chase a career path where Claude can't one-shot this kind of dumb stuff

[AussieWog93]

It's not my work. I'm not even a full time dev any more.

But the client's problem was solved, and they're happy.

This is a genuinely useful thing. You don't need to shit all over it.

[Glohrischi]

Thats quite a surprisingly arrogant take.

CRUD applications and converting business requirements into code is the thing software developers do to 99% day in day out.

[kamaal]

If you break down a complicated coding problem in smaller parts, it could be any problem.

You will see its basically a very reusable part thats already done uncountable times else where.

People who think they do something so special and novel that it just can't be done by non-human, struggle with breaking down a problem in smaller parts.

Even if you do have such novel problems, its not like every single day, every single bit of work you do is like that.

[emilsoman]

Curious, what's the career path you'd chase? Can you give examples of some work that you think Claude will never be able to one-shot?

[AussieWog93]

Oddly enough switched from software to selling retro games online.

Made ridiculous bank during 2019-2023, lost money 2024-2025 (I wasn't doing proper accounting at that stage, so it took a while to really internalise that the market wasn't insane anymore), looks like we'll make a decent-ish profit in 2025-2026 after pivoting the business model. Some regrets but it's possible staying in software could have been just as turbulent.

Funnily enough we're finally at the stage where I can launching my SaaS side-hustle which I've been sitting on for the past year and a half, so that could end up back in software again soon.

I would never say never, since I don't know what Claude would look like in 5 years' time, but there's plenty it can't do at the moment.

To give a concrete example, I don't let it make sweeping changes to the main "business logic" of my SaaS. Not because it's necessarily wrong but because I can't easily verify it. But I'll let it rip on peripheral stuff, or co-work with it.

[y0eswddl]

I'm beginning to get the sense that Sturgeon's Law is at play here and the non-crap 10% of us are arguing with the 90% for whom LLM's shitty output is actually better than what they could do on their own.

I've been lucky enough to work at places with majority intelligent engineers with similar tastes on quality to my own... but it seems to be that's not the norm or the case everywhere.

and it's the 90% that's most vocal. Sturgeon and D-K seen to go hand-in-hand.

[mwigdahl]

“Am I out of touch?”

“No, it’s the children who are wrong.”

[jaccola]

The obvious pushback to all of the slop is: coding was never hard. Learning resources were abundant and free.

If these people had a burning desire to build things prior to LLMs and couldn’t put in the effort to learn to build them (which is also fun!) then why would they ever put the effort into anything to understand it and make it good??

[layer8]

> coding was never hard. Learning resources were abundant and free.

Just a nitpick regarding “never”: Learning resources weren’t abundant and free 25 years ago, that’s a more recent development.

[skydhash]

Maybe in some parts of the world (including mine). But we haven’t have a lot of computers either. But 25 years ago, there was a lot of textbooks and computers editors like O’Reilly already active. I had the C programming language book (not 25 years ago, but the book is older than that) and you could learn a lot with that one book and codeblocks. Same thing with “The Go Programming Language”, “Learning Perl”, and “Programming Clojure”. You only need one book to get very decent.

[layer8]

True, but books weren’t free. I spent quite some money buying those.

[viking123]

Before steam was already full of games no one cares about or no one plays, like 80% won't make over 20 bucks.

It will be just garbage on top of garbage.

[HPsquared]

The basic "language translation" part was never hard, but design is hard.

[jmcodes]

What would you consider a "hard" problem?

[hexasquid]

Extrapolating the final position of the goalposts

[HPsquared]

That's a halting problem, I think.

[TeMPOraL]

And Claude can actually tackle it the same way as humans do - here in the real world, where we don't have time to let some nonsense like "mathematically proven to be unsolvable" to stand between us and our goals: it can eyeball the code and give a good enough guess.

[goatlover]

It can certainly get in a loop burning up tokens before deciding to exit some time later.

[hollowturtle]

I don't know how to define hard problems.

All I know is that we have a gigantic amount of tech debt we accumulated on the web chasing the next web framework built on top of tons of abstractions with very disappointing native web apis that shouldn't be taken seriously nor the w3c who specified them.

And when an Agent it's capable of gluing together a web app with some crud backend with a very rounded corners UI, that solves nothing for end users, we call them capable. These are not hard problems

[squidbeak]

You insist that AI needs to be able to tackle hard problems, but can't say what qualifies as a hard problem. Can you see the problem with that? If you don't know what a hard problem looks like, how do you know the models can't tackle them?

[skydhash]

It’s that it’s to able to tackle hard problems really. It’s because you have to give it the solution, and the patterns to follow, and then monitor it because it will go down weird paths.

If you’ve ever work directly with a user, you know how vague change requests can be. Try writing some vague prompts like that to the agent and see if it can solve them.

For some, writing down a (good?) specs and handing it to an agent is not very productive. Because by then, they already have an idea of the solution and can use the editor to have it done.

[simonw]

Give me a "hard problem" and I'll give you a Codex or Claude Code transcript showing how I'd use them to tackle it.

[johndough]

I've got a few. Pick whichever you like.

Factor 135066410865995223349603216278805969938881475605667027524485143851526510604859533833940287150571909441798207282164471551373680419703964191743046496589274256239341020864383202110372958725762358509643110564073501508187510676594629205563685529475213500852879416377328533906109750544334999811150056977236890927563 in less than 24 hours.

Come up with a way to sample from LLMs such that they can tell funny jokes. The jokes should not be recited jokes from elsewhere.

Implement a CUDA kernel that achieves optimal efficiency for PyTorch-like conv2d for "reasonable" shapes/strides/dilations/groups. (This task is the closest to being solved by LLMs, but they usually get stuck somewhere doing stupid things instead of considering more advanced optimization methods and still need a human to push them along).

[simonw]

I should have said "hard, but realistic".

[johndough]

Fair enough. I am optimistic that 2 and 3 will be doable eventually.