3 versions... Yikes. But the output, while messy, is still impressive, and makes me wonder how much you could trim that down by giving more comprehensive instructions.
You can test that out. Ollama does allow you to run open source models at home. I've been playing around with a bit lately and have been really enjoying it.
On my to do list is two models running at once and building a middle layer for them to interact.
One of my fun experiments recently has been putting ChatGPT in conversation mode when I go for a walk. I recently had a 45 minute conversation where "we" fleshed out a multi-agent platform. I think a key is that you need to give each agent an "inner conversation" and criteria for when output from it gets copied to the other agents and the main chat, coupled with a process to regularly compact. I intend to set up a test system I want to run continuously, and given I enjoy working on compilers maybe I'll see how much cheaper you can do something like what OP did if you orchestrator a few agents with domain knowledge in specific areas.
I think I'd want to test a state of the art model, but it'd be fascinating to see how far you can get with Ollama as well - especially whether you can compensate for less smarts by just giving it far more runtime than I could afford with e.g. Claude.
I'm sorry but substituting keywords does not make a new language.
You waited 3 months for a list of arbitrarily selected find/replace terms.
> how much did this cost - 1/4th of a San Francisco software engineer's monthly salary
Man, I really hope this isn't true. Spending 1/4 of an SF software engineer's monthly salary to set the planet on fire for a list of arbitrary and phenomenally-googleable find/replace terms demonstrates only bad qualities.
The language is the least interesting part of this. What is interesting is if it was truly built with a simple loop and without manual involvement to produce a working compiler (albeit with fairly cursed code... 5.5kloc for a parser for a grammar that appears this simple is not great)
Sadly, the linked page glosses over the specifics of the setup and instead focused on the less interesting parts..
Yea I’d expect Claude to be able to one shot this and I’m pretty pessimistic about the capability of these tools. I’m just so bewildered by posts like this.
One-shot a regexp, sure. One-shot a full compiler, not a chance.
Check out the repository - it has a full compiler that produces binaries, and a bunch of additional tooling. Doing that, even if he'd asked for a compiler for an existing language with no changes, is impressive.
Readit News
[0]: https://x.com/GeoffreyHuntley/status/1965295152962097550
On my to do list is two models running at once and building a middle layer for them to interact.
I think I'd want to test a state of the art model, but it'd be fascinating to see how far you can get with Ollama as well - especially whether you can compensate for less smarts by just giving it far more runtime than I could afford with e.g. Claude.
You waited 3 months for a list of arbitrarily selected find/replace terms.
> how much did this cost - 1/4th of a San Francisco software engineer's monthly salary
Man, I really hope this isn't true. Spending 1/4 of an SF software engineer's monthly salary to set the planet on fire for a list of arbitrary and phenomenally-googleable find/replace terms demonstrates only bad qualities.
Sadly, the linked page glosses over the specifics of the setup and instead focused on the less interesting parts..
Check out the repository - it has a full compiler that produces binaries, and a bunch of additional tooling. Doing that, even if he'd asked for a compiler for an existing language with no changes, is impressive.
Deleted Comment
> Hey, can you make me a programming language like Golang but all the lexical keywords are swapped so they're Gen Z slang?
If "struct" is "squad"
Why is it "struct" in his example?