Nobody knows how to build with AI yet • Hacker News Board

Re: Nobody knows how to build with AI yet

#391

Post by CamperBob2 » Sun, Jul 20, 2025, 6:15 PM UTC

swat535 wrote: post #353

chamomeal wrote: post #167

Earlier quoted context omitted.

I also am enjoying LLMs, but I get no joy out of just prompting them again and again. I get so incredibly bored, with a little side of anxiety that I don’t really know how my program works. I’ll probably get over it, but I’ve been realizing how much fun I get out building something as opposed to just having be built . I used to think all I cared about was results, and now I know that’s not true, so that’s fun! Of cou…

My biggest problem with working LLMs is that they don't understand negatives and they also fail to remember their previous instructions somehow. For example: If I tell it to not use X, it will do X. When I point it out, it fixes it. Then a few prompts later, it will use X again. Another issue is the hallucinations. Even if you provide it the entire schema (I did this for a toy app I was working with), it kept on maki…

I've tried many models, Sonnet is the better one at coding, 3.7 at least, I am not impressed with 4.

If Sonnet 3.7 is the best you've found, then no, you haven't tried many models. At least not lately.

For coding, I'd suggest Gemini 2.5 Pro, o3-mini-high, or Opus 4. I've heard good things about Grok 4 as well, so if you're OK with that whole scene and the guy who runs it, maybe give it a shot.

If you have already done so and still think Sonnet 3.7 is better than any of them, then the most likely explanation is that you got incredibly lucky with Claude and incredibly unlucky with the others. LLMs aren't parrots, but they are definitely stochastic.

Re: Nobody knows how to build with AI yet

#392

Post by devin » Sun, Jul 20, 2025, 6:23 PM UTC

tom_m wrote: post #381

Why are we counting the number of documents? It doesn't matter. What matters is putting together a plan and being able to articulate what you want. Then review and adjust and prompt again. You have to know how software gets built and works. You can't just expect to get it right without a decent understanding of software architecture and product design. This is something that's actually very hard. I'm coming to grips…

> The most wonderful thing I see about AI is that it should help people focus on these things.

Unfortunately I’ve been around this industry long enough to know that this is not in fact what is going to happen. We will be driven by greedy people with small minds to produce faster rather build correct systems, and the people who will pay will be users and consumers.

Re: Nobody knows how to build with AI yet

#393

Post by LtWorf » Sun, Jul 20, 2025, 6:33 PM UTC

actionfromafar wrote: post #364

LtWorf wrote: post #279

Earlier quoted context omitted.

What makes people a lead architect in my experience is an abnormal amount of arrogance and no capability to admit mistakes.

That just gives the title. To be really successful, they need to let someone else, knowledgeable, actually make the architecture decisions.

Yeah that's the actual senior developers who just ignore everything the architect architects.

Re: Nobody knows how to build with AI yet

#394

Post by CamperBob2 » Sun, Jul 20, 2025, 7:15 PM UTC

s_ting765 wrote: post #334

NitpickLawyer wrote: post #295

Earlier quoted context omitted.

> Have you also heard about calculators? Yup, my mom used to say "you need to be able to do it without a calculator, because in life you won't always have a calculator with you"... Well, guess what mom :) But on a serious note, what I'm trying to say (perhaps poorly worded) is that this is a typical thing older generations say about younger ones. They'll be lost without x and y. They won't be able to do x because the…

Junior engineers will be lost if they don't take the time to read the code generated by the LLM and really understand it. This is an objective truth. It has nothing to do with boomer takes.

Funny, that's what I said, as an experienced assembly hacker, when somebody first showed me a C compiler.

People who "take the time to really understand the code" will rapidly be outcompeted by people who don't. You don't like that, I don't like that, but guess what: nobody cares.

I suppose we'll get over it, eventually, just like last time.

Re: Nobody knows how to build with AI yet

#395

Post by CamperBob2 » Sun, Jul 20, 2025, 7:41 PM UTC

stefan_ wrote: post #128

ikerino wrote: post #78

Earlier quoted context omitted.

Hot take: Junior devs are going to be the ones who "know how to build with AI" better than current seniors. They are entering the job market with sensibilities for a higher-level of abstraction. They will be the first generation of devs that went through high-school + college building with AI.

Where did they learn sensibility for higher-level of abstraction? AI is the opposite, it will do what you prompt and never stop to tell you its a terrible idea, you will have to learn yourself all the way down into the details that the big picture it chose for you was faulty from the start. Convert some convoluted bash script to run on Windows because thats what the office people run? Get strapped in for the AI Power…

AI is the opposite, it will do what you prompt and never stop to tell you its a terrible idea

That's not true at all, and hasn't been for a while. When using LLMs to tackle an unfamiliar problem, I always start by asking for a comparative review of possible strategies.

In other words, I don't tell it, "Provide a C++ class that implements a 12-layer ABC model that does XYZ," I ask it, "What ML techniques are considered most effective for tasks similar to XYZ?" and drill down from there. I very frequently see answers like, "That's not a good fit for your requirements for reasons 1, 2, and 3. Consider UVW instead." Usually it's good advice.

At the same time I will typically carry on the same conversation with other competing models, and that can really help avoid wasting time on faulty assumptions and terrible ideas.

Re: Nobody knows how to build with AI yet

#396

Post by pksunkara » Sun, Jul 20, 2025, 8:36 PM UTC

dangero wrote: post #264

I want claude code on my phone running in a cloud vm so I can give feedback out on a trail somewhere and continue my three hour hike or bike ride with my family.

I am working on something for that at https://automa.app. Let me know if you are interested in trying it out.

Re: Nobody knows how to build with AI yet

#397

Post by wiremine » Sun, Jul 20, 2025, 8:44 PM UTC

gronglo wrote: post #220

wiremine wrote: post #138

I've been experimenting with model-based development lately, and this resonated strongly with me. The section "What Even Is Programming Anymore?" hit on a lot of the thoughts and feels I've been going through. I'm using all my 25+ years of experience and CS training, but it's _not_ programming per se. I feel like we're entering an era where we're piloting a set of tools, not hand crafting code. I think a lot of peopl…

> I think a lot of people (who love crafting) will be leaving the industry in the next 5 years, for better or worse. I think you're spot on. It was once necessary to acquire knowledge in order to acquire productivity. This made knowledge valuable and worth attaining. Now, with LLMs, we we can skip the middle man and go straight to the acquisition of productivity. I'd call it the democratisation of knowledge, but it's…

> knowledge just isn't needed anymore

I would argue we still need the knowledge: the principles aren't changing, and they are needed to be truly productive in certain things. But the application of those principles _are_ changing.

Re: Nobody knows how to build with AI yet

#398

Post by s_ting765 » Sun, Jul 20, 2025, 8:45 PM UTC

CamperBob2 wrote: post #394

s_ting765 wrote: post #334

Earlier quoted context omitted.

Junior engineers will be lost if they don't take the time to read the code generated by the LLM and really understand it. This is an objective truth. It has nothing to do with boomer takes.

Funny, that's what I said, as an experienced assembly hacker, when somebody first showed me a C compiler. People who "take the time to really understand the code" will rapidly be outcompeted by people who don't. You don't like that, I don't like that, but guess what: nobody cares. I suppose we'll get over it, eventually, just like last time.

LLMs are not compilers. They can't be deterministic. A better comparison is an autocorrect on steroids.

And I don't think there's anything to get over about them. They are useful but people elevate their significance too much over what they actually are.

Re: Nobody knows how to build with AI yet

#399

Post by CamperBob2 » Sun, Jul 20, 2025, 10:27 PM UTC

s_ting765 wrote: post #398

CamperBob2 wrote: post #394

Earlier quoted context omitted.

Funny, that's what I said, as an experienced assembly hacker, when somebody first showed me a C compiler. People who "take the time to really understand the code" will rapidly be outcompeted by people who don't. You don't like that, I don't like that, but guess what: nobody cares. I suppose we'll get over it, eventually, just like last time.

LLMs are not compilers. They can't be deterministic. A better comparison is an autocorrect on steroids. And I don't think there's anything to get over about them. They are useful but people elevate their significance too much over what they actually are.

An unhealthy attachment to determinism will turn out to be a career-limiting hangup, I suspect. You already lack insight into how 100% of the code in your project works, unless you only work on trivial projects. Did you think that state of affairs was going to get better with time? As usual, TDD covers a multitude of sins.

As for "autocorrect," let us know when your "autocorrect" takes gold at the International Math Olympiad, with or without steroids.

Re: Nobody knows how to build with AI yet

#400

Post by throw234234234 » Sun, Jul 20, 2025, 11:01 PM UTC

doug_durham wrote: post #278

yoyohello13 wrote: post #223

Earlier quoted context omitted.

So like, why would I pay someone to work like this? Or why would I pay for software at all when I could just do this myself?

Because you can't do this yourself. You don't have the decades of experience to know how to ask the questions and when to steer the tool into a different direction.

Why not? As the AI people say "you can just do things". If there is little consequence to getting it wrong as a parent poster puts it why not? You can learn over time like any other skill.

I want to be wrong (it will affect me and my family personally) - but there is a reason every AI proponent talks about coding and making "coding redundant". For most jobs/industries software is a compliment (e.g. Product Owner, BA, etc) unless that is your main skill in which case it is your main service you are selling. Most roles want to turn software into a commodity or the typical business PM word "resource" that they can acquire as they need - the dream of most business roles (e.g. Project Managers, BA's, etc). It sadly also seems to be low hanging fruit of LLM's; doesn't mean there isn't other aspects to the job of course but coding is becoming "less special" with these technologies especially with common tech and use cases.