We've started to be an early user in December and have since adopted it in a brown field codebase. I'd describe this as Lovable 2.0 or vibe coding 2.0. The p0 workflow allowed us to delegate medium to complex full features to Claude Code while staying in lane with our standards. It allows us to go from idea to fully working prototype and PR within half and hour to hour, most of the time fully hands off. PRs still need to be reviewed for production. However p0 allowed to drastically improve per engineer velocity, AI code quality and iterate faster with working prototypes and refinements.

Compared to Claude Code directly, which we also use heavily, p0 keeps very strong coherence from user story, spec planning, architecting, engineering and QA - across many several agents and subagents. Breaking down the work into sequential and parallel task. With Claude Code alone this would be usually requiring lots of hand holding, or be be only partially focussed, rest lost in the woods. Also, we attempted to replicate some of p0 ideas with home grown software dev personas and workflows which fell apart. I think the strong point of p0 is that they really nailed the decomposition and software dev cycle with agents.

Really recommend to try, and at the very minimum you get to make your codebase agent ready if you haven't already.

Anyone aware of a practical how to on implementing a data flywheel for fine-tuning (improving the model with user feedback)?

Not seen a great explainer on this yet.

You'd either need access to the model weights or a fine-tuning API.

Then depending on which fine-tuning approach you want to use, the user data you need to collect will be different: RLHF requires multiple outputs to a single query vs instruction fine-tuning where you need great input-output pairs to train on. You could ask the user's feedback after running the LLM to pick out good training data.

Thanks for sharing. Why is the training dataset that contains instructions and output wrapped by another enclosing prompt (https://github.com/minosvasilias/godot-dodo/blob/f62b90a4622...)

Why does this even work when during inference this wrapping prompt is absent? Wouldnt the model then work best against a inference prompt that follows the wrapping prompt structure, however the desired outcome is to have a model that just works without the wrapping prompt?

Edit: see reply from OP, the wrapping prompt is used for inference as well, so misunderstanding on my part

The wrapping prompt is also used during inference. (https://github.com/minosvasilias/godot-dodo/blob/f62b90a4622...) Prompting like this is useful for instruct-finetunes, and similar prompts are used by other projects like stanford-alpaca.

Thanks for the clarification, makes sense now!

It's using Google Slides[0] which then allows to export to PPTX. There may be other better options.

[0] https://developers.google.com/apps-script/guides/slides

Obvious miss, fixed.

So you corrected by ensuring to censor genesis as well?

Dupe of my own submission earlier: https://news.ycombinator.com/item?id=35066876#35070202

OP here. The purpose here of the "health" filter, which in its current state is fairly basic, is to to provide minimum protection for sensitive topics. For example, there are a larger number of health related requests and at the current state of quality/correctness of output I found it to be irresponsible to generate such content.

This may be an indication that it's at least useless, if not irresponsible, in other contexts too.

Can you let me know your prompt so I investigate on the images? It's currently using Unsplash which has a very permissible license. I need change the endpoint from Unsplash Source to their API which then allows to do proper attribution.

Thank you slides are being added depending on the mood/temperature of the model :)

Will adjust the prompt to include proper agenda and Thank you note.

You’re welcome to find it from logs — I honestly didn’t keep the prompt, and you seem to not provide the prompt back on the output screen (which would be very helpful for this purpose alone).

https://slidesgpt.com/l/TT58

OP here.

Tried the same with "Create a slide deck on the future of AI - use the iconic style and tone of an Apple Keynote" it generated https://slidesgpt.com/l/NqZ4

Looks like an area to improve on to support different tones and styles

OP here. The sweet spot seems to get off the ground quickly for now. I'm currently prototyping refinements/iterations so once the initial deck is created you can directly add your feedback to redo parts of the deck and gradually improve/steer the output. Thanks for testing and feedback!