Live data from Hacker News

AI coding at home without going broke

stephen.bochinski.dev

281–290 of 321 posts

Re: AI coding at home without going broke

#281
post #48
post #3

> The first is to self host. You buy the machine, run open source models locally, and pay nothing per token after that. Power is not free. What I’ve found is that you’re basically paying a premium for privacy, and that’s worth it for me.

Luckily I needed a new laptop and I bought an M1 Max secondhand from a friend quite cheaply because it was fast enough to recompile something else I am interested in. So for me, there is no additional hardware cost; it was acquired in replacement. I run the AI models at home on this kit because I want to; I'll use openrouter if I need to. I accept the economics of this article are right. But I feel so incredibly sad…

So basically you don't have a life outside of the job?

> now really unfixable because I am a bit too old

How old is a bit too old? I know 50+ colleagues doing sports and traveling just fine.

Re: AI coding at home without going broke

#282

Earlier quoted context omitted.

I've found that adding "Make no mistakes." to my prompt usually helps with this kind of problem...

Or if the code is really important, sometimes even “please make no mistakes” is necessary.

[dead]

Re: AI coding at home without going broke

#283
The big one that's missing as cost cutter. If you are a programmer, stay as close as you can to the code. Define the interfaces yourself and the core logic. You will know exactly what changes you want and know what tier of model to use.

Re: AI coding at home without going broke

#284

Earlier quoted context omitted.

recommend converting to https://github.com/apple/container

Linux VM doesn't run native macOS toolchain and requires copying files back and forth

If you don't want to do that, don't use a VM. I like nono:

https://github.com/always-further/nono

Re: AI coding at home without going broke

#285

Earlier quoted context omitted.

Footprint's comment is correct. I go directly to Deepseek's platform API which they linked. There's no image generation but you get access to Deepseek V4 Flash and Deepseek V4 Pro, both of which are very capable for general text based tasks and programming. Flash is insanely cheap for how good it is ($0.14 per 1M input tokens vs $15 with Claude 4.7). V4 Pro I would put somewhere in the range of 80 to 90% as good as O…

This is awesome, thank you for posting this. Are you using pi, Hermes, or another harness? I’d love to hear more about your workflow.

I've been using [OpenCode](https://opencode.ai/) - I find it works quite well and has things like web search and build/plan modes built in. I had to modify the settings though (On Linux at `~/.config/opencode/opencode.json`) to stop it from just modifying files without first asking for permission, which I didn't like. I like being able to read the changes my AI agent suggests before the files are modified.

Re: AI coding at home without going broke

#286
post #276
post #2

I find just going via Deepseek's platform API directly, using their V4 flash model, and hooking into a harness like Opencode more than acceptable. Think I've spent maybe $10 over a couple of weeks. I did explore self-hosting models but hardware right now is just too expensive.

I believe opencode go but only using deepseek flash would last you longer. (Equivalent to $65 in tokens but it's a monthly payment so you have to be using it up or deepseek direct will be cheaper) First month is $5, later $10. Cancel any time. You can keep getting the deal with a new email.

Interesting. Thanks for letting me know. I will investigate that if I end up finding the API too expensive.

Re: AI coding at home without going broke

#287
post #48

Earlier quoted context omitted.

Luckily I needed a new laptop and I bought an M1 Max secondhand from a friend quite cheaply because it was fast enough to recompile something else I am interested in. So for me, there is no additional hardware cost; it was acquired in replacement. I run the AI models at home on this kit because I want to; I'll use openrouter if I need to. I accept the economics of this article are right. But I feel so incredibly sad…

So basically you don't have a life outside of the job? > now really unfixable because I am a bit too old How old is a bit too old? I know 50+ colleagues doing sports and traveling just fine.

> So basically you don't have a life outside of the job?

I very much did. A big social life. I may have had more of a social life than most, in fact; I have been so lucky.

But that doesn't change that I am the kind of person who drew a lot of self-worth from being able to use my skills to help people, which was actually what helped me find that life in the first place.

As I mentioned elsewhere I have been dealing with really profound burnout for a couple of years. It is extremely difficult. It has made it distressing to try to cope with busy social environments; I am not hiding but life has changed and other people's lives move on. (Including other freelancers you work with.)

Without a sense of engagement from what I do for a living, I am left with a lot less of a life. And the world of tech has changed in so many ways in just the time I have been trying to recover that it feels difficult to find a place again. It is taking an enormous amount of mental energy to catch up when I have sometimes only been able to focus for about an hour in any three days.

Lately things have been better, and while I am AI-cynical I have been enjoying digging into a new topic and working out what I think, but doing this past the age of fifty when you're burned out is hard work.

> How old is a bit too old?

I was talking about rethinking priorities and really seeking a family life for myself there, and the answer is that I am enough over fifty that there is no fair way to approach that, given my current mental health.

Some things, I'm afraid, do just one day stop being possible. You may not get a do-over after, say, forty-five.

Re: AI coding at home without going broke

#288
post #133

Earlier quoted context omitted.

Patches to existing sizable codebases and reverse engineering binaries both can run a long time and use a lot of tokens without wandering off into the weeds.

Claude allows you to reverse engineer binaries now? That's pretty cool. I'm quite surprised to hear that, I thought it was one of their guardrails. Most of the reverse engineering projects I've seen seem to rely on Chinese models.

The guardrails are probably sensitive to what the target is and how you frame it. If it's "I want to help preserve this old video game by decompiling it" then ok, if it's "decompile this industrial control software so I can do a terrorism" then I'd expect it to refuse.

Re: AI coding at home without going broke

#290
post #100

I feel like I must have plateued and don't know what to do next to level up. I'm currently on the $100/month codex plan and it seems fine using 5.5-xhigh all the time. I think of what to do next, have a chat session to determine exactly what to ask for up to the point of being ready to implement, and then codex churns on a commit-sized task whereupon I briefly check it on my local dev server. If necessary I ask for a…

I have downgraded my Claude to the $20 one, and basically only use it for the web chat right now. For coding, I use DeepSeek @API Rates configured in Claude Code. I have spent around $4.8 for 320,000,000 tokens. I always felt like i was not using Claude plan, that i had to have the LLM working on something all the time to justify the price. Now with DeepSeek i don't think about it anymore. I don't feel bad when not u…

If I’m reading right you used to pay more for Claude but now deepseek has replaced that higher tier subscription. Do you mind my asking what you were paying before?
Post reply on HN