It's a fascinating paradigm – as though it dropped here out of some future.
However, I'd never use it. Reason: having everything dynamically created JIT means there's a trade-off somewhere else; in this case, it's energy consumption. I'd rather "cache" applications (ie. have them as installable software, which has been built just once) than endlessly recreate software from scratch each time, with all the energy usage that entails.
One critique is that I think it might be more efficient to just make use of existing software for some of these things rather than generate them on the fly (for something for example like a text editor), but I guess that will depend on how costly it is to run AI programs in the future (if it becomes much cheaper, then I could see your concept being taken up by more people)
Yes that is definitely still gonna be necessary and it's why I build on a store where people can upload their apps/agents. There is a loose community section and in the backlog I have work around offering an sdk and official store for established companies to ship pneuma apps.
For this to be practical you really need to build things iteratively and asking the user questions, rather than attempting at one shot half assed ideas
Also most people are better explaining details with voice rather than text.
Voice is a great idea! On the backlog.
Also pneuma already supports iteration/edits. You can say "edit my pong game so the paddles are pink" and it's smart enough to do it.
I think this is the way things are ultimately going (software as ephemerally generated based on your preferences), but at the moment, I can't see myself using these generated apps. Maybe it's just an aesthetic thing—have you experimented with personalized theming?
I've built a list of common gotchas in the generation prompts.
Also if the compilation fails it falls back to opus with the error message and code and can try again twice.
Thanks! Playbit is a cool project, I've followed it. The key difference with Pneuma is that the programs themselves are AI-generated at runtime. There's no toolchain the user interacts with, no editor, no compiler invocation. You describe what you want and the system handles everything from code generation through compilation to execution in one pass. Yansu is interesting too but it's a web app builder. Pneuma runs native GPU-rendered agents in WASM sandboxes on the desktop, so you can build things like real-time games and data visualizers with keyboard/mouse input, not just web UIs. The closest analogy is probably a microkernel where every userspace process is conjured by an LLM instead of installed from a package manager.
That makes sense—I think it's really interesting to see the possibilities of new software created purely by AI! I'm happy to offer feedback, my email is micah.blachman@gmail.com
However, I'd never use it. Reason: having everything dynamically created JIT means there's a trade-off somewhere else; in this case, it's energy consumption. I'd rather "cache" applications (ie. have them as installable software, which has been built just once) than endlessly recreate software from scratch each time, with all the energy usage that entails.
Does that mean it's not actually an OS?
For this to be practical you really need to build things iteratively and asking the user questions, rather than attempting at one shot half assed ideas
Also most people are better explaining details with voice rather than text.
Pneuma reminds me a little bit of Yansu AI, a project I saw recently on HN that proactively builds apps with AI (https://yansu.app/)