Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

After months of coding with LLMs, I'm going back to using my brain

albertofortin.com

Post reply on HN
Print view View original

41–50 of 229 posts

Re: After months of coding with LLMs, I'm going back to using my brain

#41

Post by picklesman »

qwertox wrote: post #27

Go to AI Studio, select Gemini Pro, give it your code, or describe a problem you want to solve, and then tell it that you want to --> discuss This is the best way to get Gemini to be a really good assistant, unless you want to add System Instructions which precisely describe how it should behave. Because if you just say it should solve some problem for you, it eagerly will generate a lot of code for you, or add a lot…

Yeah my current approach is to generate a plan with Gemini Pro, with plenty of back and forths, and then have it write the plan to a markdown file. Afterwards I get either it or another model to follow the plan step by step. Without doing this the results are questionable and often require going back and fixing a lot.

Re: After months of coding with LLMs, I'm going back to using my brain

#42

Post by sublinear »

jclardy wrote: post #8

I don't get the whole "all-in" mentality around LLMs. I'm an iOS dev by trade, I continue to do that as I always have. The difference now is I'll use an LLM to quickly generate a one-off view based on a design. This isn't a core view of an app, the core functionality, or really anything of importance. It's a view that promotes a new feature, or how to install widgets, or random things. This would normally take me 30-…

> I also use it for building things like app landing pages.

This is a reasonable usage of LLMs up to a certain point, and especially if you're in full control of all the requirements as the dev. If you don't mind missing details related to sales and marketing such as SEO and analytics, I think those are not really "landing pages", but rather just basic web pages.

> I hate web development, and LLMs are pretty good at it because I'd guess that is 90% of their training data related to software development.

Your previous sentence does not support this at all since web development is a much more broad topic than your perception of landing pages. Anything can be a web app, so most things are nowadays.

Re: After months of coding with LLMs, I'm going back to using my brain

#43

Post by rco8786 »

a7fort wrote: post #39
reconnecting wrote: post #5

Off-topic: may I ask, why PHP + MySQL is no longer considered suitable to you? It's hard to say without specifics, but simply upgrading from MySQL to PostgreSQL without rewriting the PHP codebase in Go might resolve most of the potential issues.

It's two separate issues: 1) I do a lot of scraping, and Go concurrency + Colly has better performance 2) My DB size is exploding and I have limited budget, and it looks like CH is so much better at compressing data. I recently did a test and for the same table with same exact data, MySQL was using 11GB, ClickHouse 500MB

> I recently did a test and for the same table with same exact data, MySQL was using 11GB, ClickHouse 500MB

That's pretty impressive

Re: After months of coding with LLMs, I'm going back to using my brain

#44

Post by Molitor5901 »

I love using LLMs with my debugging but also helping me to get started. Grok has been helpful in giving it a rubric and having it come up with the framework and headers, which saves some time. Unfortunately in my humble and limited experience, asking it to then improve upon that code is where it runs into mistakes. If you give it a perfect rubric of what you want, I think an LLM can give you some very good code, but it can't think ..beyond that like a human can.

Something I do very much love about LLMs is that I can feed it my server logs and get back an analysis of the latest intrusion attempts, etc. That has taught me so much on its own.

Re: After months of coding with LLMs, I'm going back to using my brain

#45

Post by 3dsnano »

This perspective can only come from someone who started programming in the pre-LLM era. Back in those days the code didn’t write itself and you had to think really hard about shit and bang your head in the wall. This literally requires strong and coherent mental models, because there’s no other way… there were no shortcuts.

Post-LLM era, you can use a shortcut to get a representation of what something could be. However I see a lot of folks shipping that representation straight to prod. By comparison, the mental models created are weaker and have less integrity. You might be able to feel that something is off, but you lack the faculties to express or explain why.

Re: After months of coding with LLMs, I'm going back to using my brain

#46

Post by reconnecting »

a7fort wrote: post #39
reconnecting wrote: post #5

Off-topic: may I ask, why PHP + MySQL is no longer considered suitable to you? It's hard to say without specifics, but simply upgrading from MySQL to PostgreSQL without rewriting the PHP codebase in Go might resolve most of the potential issues.

It's two separate issues: 1) I do a lot of scraping, and Go concurrency + Colly has better performance 2) My DB size is exploding and I have limited budget, and it looks like CH is so much better at compressing data. I recently did a test and for the same table with same exact data, MySQL was using 11GB, ClickHouse 500MB

The actual makes sense. I thought you had a PHP application that you just decided to rewrite to Go.

Re: After months of coding with LLMs, I'm going back to using my brain

#47

Post by casey2 »

I promise you do not need to "use your brain"

Use the tool when it makes sense or when someone shows you how to use it more effectively. This is exactly like the calculator "ruining people's ability to do arithmetic" when the vast majority of the population has been innumerate for hundreds of thousands of years up til the IR where suddenly dead white european nobility are cool.

There is nothing fun nor interesting about long division as well as software development.

If LLMs don't work for your usecase (yet) then of course you have to stick with the old method, but the "I could have written this script myself, I can feel my brain getting slower" spiel is dreadfully boring.

Re: After months of coding with LLMs, I'm going back to using my brain

#48

Post by jdiff »

VMG wrote: post #11

I get it and I see the same problems as the author. I'm working on a few toy projects and I am using LLM for 90% of it. The result is 10x faster than if I coded it "by hand", but the architecture is worse and somewhat alien. I'm still keeping at it, because I'm convinced that LLM driven code is where things are headed, inevitably. These tools are just crazy powerful, but we will have to learn how to use them in a way…

At least currently, the only use pattern that can withstand complex codebases is as advanced speech-to-text, but without the speech if that makes sense. The main issue with that is phrasing things in English is often far more verbose, so without the speech, it's very often faster to just do it by hand.

Re: After months of coding with LLMs, I'm going back to using my brain

#49

Post by emseetech »

mritchie712 wrote: post #37
jclardy wrote: post #8

I don't get the whole "all-in" mentality around LLMs. I'm an iOS dev by trade, I continue to do that as I always have. The difference now is I'll use an LLM to quickly generate a one-off view based on a design. This isn't a core view of an app, the core functionality, or really anything of importance. It's a view that promotes a new feature, or how to install widgets, or random things. This would normally take me 30-…

How useful the various tools will be depends on the person and the problem. Take two hypothetical people working on different problems and consider if, for example, Cursor would be useful. IF you're a: * 10 year python dev * work almost entirely on a very large, complex python code base * have a pycharm IDE fine tuned over many years to work perfectly on that code base * have very low tolerance for bugs (stable produ…

Agreed but the 1 year JS dev should know they're making a deal with the devil in terms of building their skillset long term.

Re: After months of coding with LLMs, I'm going back to using my brain

#50

Post by andy99 »

jclardy wrote: post #8

I don't get the whole "all-in" mentality around LLMs. I'm an iOS dev by trade, I continue to do that as I always have. The difference now is I'll use an LLM to quickly generate a one-off view based on a design. This isn't a core view of an app, the core functionality, or really anything of importance. It's a view that promotes a new feature, or how to install widgets, or random things. This would normally take me 30-…

Overshooting the capabilities of LLMs is pretty natural when you're exploring them. I've been using them to partially replace stack overflow or get short snippets of code for ~2 years. When Claude code came out, I gave it increased responsibility until I made a mess with it, and now I understand where it doesn't work and am back to using LLMs more for ideas and advice. I think this arc is pretty common.
Post reply on HN