Is this seriously so? Have you never seen anything helpful from an LLM? That seems such a black and white statement that I get confused.

I am conservative regarding AI driven coding but I still see tremendous value.

It makes me want to ask you: do you ever see helpful things from your colleagues at all?

> Is this seriously so? Have you never seen anything helpful from an LLM?

No, not at all. I may be using it wrong.

I put in "write me a library that decodes network packets in <format I'm working with>" and it had no idea where to start.

What part of it is it supposed to do? I don't want to do any more typing than I have to.

You're right, you are using it wrong. An LLM can read code faster than you can, write code faster than you can, and knows more things than you do. By "you" I mean you, me, and anyone with a biological brain.

Where LLMs are behind humans is depth of insight. Doing anything non-trivial requires insight.

The key to effectively using LLMs is to provide the insight yourself, then let the LLM do the grunt work. Kind of like paint by numbers. In your case, I would recommend some combination of defining the API of the library you want yourself manually, thinking through how you would implement it and writing down the broad strokes of the process for the LLM, and collecting reference materials like a format spec, any docs, the code that's creating these packets, and so on.

> An LLM can read code faster than you can, write code faster than you can, and knows more things than you do.

I don't agree. It can't write code at all, it can only copy things it's already seen. But, if that is true, why can't it solve my problem?

> The key to effectively using LLMs is to provide the insight yourself, then let the LLM do the grunt work

Okay, so how do I do that? Remember, I want to do ZERO TYPING. I do not want to type a single character that is not code. I already know what I want the code to do, I just want it typed in.

I just don't think AI can ever solve a problem I have.

No such thing as a free lunch. Zero typing is an unrealistic and unachievable goal

Well, if you ask it to write a library at the start, it's likely it will not do that well. Start small, spoon feed some examples.

If you have to put in this much effort, why not just write it yourself?

When you write a library the first step is always designing it. LLMs dont get rid of that step, they get rid of the next step where you implement your design.

They also added an additional step where you have to explain your design using vague natural language.

Is this really "additional"? do you not do design docs/adrs/rfcs etc and talk about them with your team? do you take any notes or write out your design/plan in some way even for yourself?

Why can't you just pass any of those to an AI?

If I'm writing a library to work with a binary format, there is very little English in my head required, let alone written English.

That is a heavily symbolic exercise. I will "read" the spec, but I will not pronounce it in literal audible English in my head (I'm a better reader than that.)

I write Haskell tho so maybe I'm biased. I do not have an inner narrative when programming ever.

I’m not part of any team, I work on my projects alone. I rarely write long-form design documents; usually I either just start coding or write very vague notes that only make sense when combined with what’s in my head.

some people suck ass at programming so they'd rather use English

So I have to do a lot of typing? Because the typing is the bit I don't want to do.

Actually writing code is the fun and easy bit.

I think one side of the issues folks are having is that combined with the mandate to use these tools, there is also an expectation or assumption that the developers will instantly get X% more productive. Like, "you must use this tool and you will be twice as productive".

Where I work there as certainly been that kind of discussions, "we need to use AI for this, because no offense but you are simply not fast enough". And this from people who do not understand software development and has never worked with it. They have only read the online stuff about 20X speeds and FOMO. (And my workplace is generally quite laid back and reasonable. I am sure many other places are much more aggressively steered.)

It will simply create it?

I am stealing that quote.

I can not claim it )) I think that Eisenhower said it first.

Many say but I don't agree. It is clearly better now but I had basically the same view on code gen-AI a year ago as I have now. It was obvious even then that LLMs were a big deal. They were really cool then and are amazing now. But some issues are undeniably still there. Maybe they are not a question of some simple quality measure, meaning they might not be solved by simply crunching more tokens with larger context.

I used Claude Code before August 2025 and it was definitely usable, although clearly more capable now. The difference is noticeable but not a completely different world, all in all, in my eyes.

I notice on a daily basis even now that it can easily lead to bloat and unnecessary complexity. We will see if it can be fixed by using even stronger models or not.

Was it Bill Gates who likened LoC to measuring airplane construction progress by weight?

Out of curiosity, did you also implement scramble support? Or just the timing stuff?

yes. claude added a suggested random scramble (if that's what you mean?), also running average of 5/12/100, local storage of past times on first iteration, my son told it to also add a button for +2s penalties and touch screen support.

Ok cool! I have not done any cubing related coding so I don't know how complicated it gets but making sure suggested scrambles are solvable etc seems like it could be non-trivial?

You just get a sequence of random moves to go from solved to scrambled, it's quite trivial.

See here if you're interested: stefansiegert.net/cube-timer

Let me know if you adapt it in any way, my son would be delighted to see open source work its magic :)

Ah of course, thank you. Defining the moves to get to the scramble makes sure it is solvable.

Good advice, thank you for the reminder.

Regarding the OP's dilemma. I am split. I enjoy both the process and the destination. With AI, the process is faster and less satisfying, but reaching the destination is satisfying in its own way, and enables certain professional ambitions.

I have always had other outlets for my "process" needs, and I believe I will spend more time on them in the future. Other hobbies. I love "artisanal coding" but that aspect was never really my job.

It's classified.