Hacker News Reader

Comment by fooster | Hacker News Reader

I get excellent results and don’t do anything like that. Basically I ask Claude to write code as I do. A small step at a time. I literally prompt it to do the next step I’d do and so on and so forth. I accept all changes immediate and then commit after every change and then review the diff. If Claude did some badness then I ask it to fix that. I typically also give references to existing code that I want it to model or functions to use.

This gives me excellent results with far less typing and time.

View on HN

Exactly this. In my view people over-think the use of Claude in many cases.

Because it's marketed as AI and it takes a while to figure out that it's really quite limited. In my opinion there's not a lot of intelligence going on, It's great at translating a requirement and giving you an approximation of what you asked for, but there isn't really any "thinking" going on.

I think when they advertise "thinking" it just does a few more iterations of giving you the closest "number in your head from the clues you've given it (requirements).

I saw someone once say that LLMs are a kind of "word calculator" and I feel that's quite a good description.

To be fair, if you turn on thinking mode on Llms and see the output their is some thinking/reasoning.

A simple example:

Prompt:Make it yellow

Think: the user wants something yellow but hasn't said what it is. Previously the user talked about creating a Button, so it must be the button but I should clarify by asking

Response: is it the button that you want yellow?

this part of LLMs is the pre-programmed human logic. The LLMs arent actually thinking, they're just going through a defined IF THEN loop based on whatever weights, if there is some ambiguity in a prompt the LLM is just programmed to prompt for more info. Its not actually thinking it needs to ask anything, its just coming back with low precision probability.

All of the recent improvements in the LLM's "thinking" have just been layers of programming on top of its statistical models. Which is why its becoming clear LLMs are really not advancing that much.

I guess it becomes philosophical, what is thinking? What I’ve notice it does all that but misses the absolute most obvious things or gets incredibly basic things wrong.

It is much better at making todo lists than it is at completing them successfully. But I appreciate having “someone to chat with” when I’m puzzling over codes or what to do next. Especially when I’m getting started on something new I don’t know a lot about.

I don't understand why the diff isn't part and parcel of the AI dialog..

In VS code it shows me a diff with every contribution.

Same with other IDEs like cursor.

But Claude Code is a command line tool.

You can use the Claude code add on and it shows diffs in vscode and you can highlight code and ask questions. But it still works in the vscode terminal like Claude code the cli. It’s quite nice actually since you can flip back and forth.

This is how I do it. Instead of writing a large amount, I just make it generate one function at a time.

I dont want it to replace me, I replace reading the docs and googling or repetitive tasks.

Its a hit or miss sometimes but I get to review every snippet.

If I would generate alot of code at once like for a full project I would get brainfuck reviewing it.

I keep my normal development flow and iterate, no waterfall

Sometimes I do OP’s approach, sometimes yours, but in all cases, writing down what you need done in detailed English gets me to a better understand of what the hell I’m even doing.

Even if I wrote the same prompts and specs and then typed everything myself, it would have already been an improvement.

I've come to the conclusion that the best use for "AI" it typing faster than me. I work at a place with a very well defined architecture so basically implementation is usually very straight forward, Claude can follow it because as you said, it's following a spec of sorts.

On the other hand, there has been quite a few moments in the last week where I'm actually starting to question if it's really faster. Some of the random mistakes can be major depending on how quickly it gets something wrong. I feel like a computer game, I need to save every time I make progress (commit my work).

I'm still on the fence about it honestly.

Wouldn’t it be faster to write the code yourself at that point ?

What’s the advantage here for you with a process like this?

In your flow you also have multiple review steps and corrections as well adds even more friction.

I can see the advantage in what parent is describing however.

Sometimes its faster I write it, sometimes its not.

I prompt like: give me go structs for this json "pasted json" and write 2 functions to save it and load it from "nosql db I use"

That basically speeds up writing glue code

The business logic I write myself

Its faster to do the business logic myself than review what the Ai did.

> give me go structs for this json "pasted json" and write 2 functions to save it and load it from "nosql db I use"

Those tools existed before LLMs and were local, fast, and most importantly free.

Why people continue to re-invent tools and workflows we already have, I don't know. Perhaps they just like to be able say "but this uses AI!"

It was an example

I only use what I need, plain and simple, and its free yup, works local too.

Llms suck at large projects, so I break it down to small simple things, what I wrote is an example of a small thing, I keep the tasks that large.

I can write my own code. I just use LLMs to speed up. I use whatever workflow suits me the most and saves me from typing much

Code generating with LLMs is a reinvention of local tools that used to generate code, yes. Its an advanced autocomplete.

But give me another tool that can generate code from natural language full of typos. I dont think your claim that it existed before llms is correct at all. Ever code generation task could be handwritten to be reusable but that takes developer hours

> give me another tool that can generate code from natural language full of typos

Give me a tool that doesn't generate code full of errors.

Reply on news.ycombinator.com