We’ve been working towards a similar goal in a different way. We still use English to talk to the LLM and python for code, but we’re working on a system where you can mix the two much more conveniently.
@jeremyphoward
-
Hybrid evolutionary algorithms and current LLM limitations
By
–
That uses a hybrid evolutionary algorithm. It’s a nice approach but doesn’t really help with the current LLM problems
-
LLMs Performance Degradation at Critical Thresholds
By
–
Yeah all LLMs suddenly become *extremely* dumb when you hit that point.
-
GraphQL OpenAPI interfaces trivially accessible through built-in API documentation
By
–
Plus of course anything that provides a graphql or openapi interface is trivially available through tools thanks to their built-in API docs.
-
Remote Server Access via Shell Scripts and Claude Code
By
–
A lot of remote servers are accessible by shell scripts and Claude Code too though, like `gh`, `ghapi`, etc 😀
-
FastTag: HTML Templating with M-Expression Syntax
By
–
FastTag *is* html. It just replaces XML syntax with m-expression syntax. It's a 1-1 direct mapping. You *must* have your tag syntax directly in your actual code, to treat it natively. Just like jsx. I don't know anything about that benchmark — that's not a lib we built.
-
When to Stop: Evaluating Startup Purpose and Viability
By
–
Oh also, you need to be doing something worth doing for a long time. If it turns out it's actually not worth doing, then you may as well stop doing it. Keeping your startup alive for no reason is pretty pointless.
-
Maintain Patience: Work Fun, Sleep Well, Stay Mission-Focused
By
–
You can avoid running out of patience by ensuring the work stays fun, working with people who are nice to work with, ensuring that you stay on-mission forever so you care about what you do, and exercising and sleeping well. Or some combination of those things.
-
Startup Funding Strategies: Raising Capital and Managing Runway
By
–
You can avoid running out of money by raising a lot at the start, or raising a lot as you go, or by controlling your costs so your money lasts a long time, or by getting revenue quickly and spending less than you make. Or some combination of those things.