why is this better than eg matplotlib?
Show HN: Microsoft releases Flint, a visualization language for AI agents
141–149 of 149 posts
Re: Show HN: Microsoft releases Flint, a visualization language for AI agents
#142Doubt this is really a language problem instead of a model-capability one. A better prompt might get you most of the way there.
Both, it's a combination of latency / reliability / expressivenss. I think when models are more powerful, we will also more naturally want to create more expressive stuff.
Re: Show HN: Microsoft releases Flint, a visualization language for AI agents
#143I think a lot of were already using Mermaid and/or Python/Matplotlib(etc) for this. What would be the advantages to Flint?
Mostly for expressiveness and reliability & cost trade-off. Flint has advantage of being an intermediate language that allow agents to generate good-looking stuff without additional refinement loops, since the compiler derives lower-level geometric constraints from semantic types.
Would especially be handy when we are building some agents that produce charts that serve end users (they want faster and more reliable experiences!).
Re: Show HN: Microsoft releases Flint, a visualization language for AI agents
#144can this be used for math viz? Im having claude teach me visualize math and it writes lots of low level canvas code just to make it look good.
no yet! that's a good direction!
Re: Show HN: Microsoft releases Flint, a visualization language for AI agents
#145I find that agents can use charts pretty well though. Maybe it would be interesting to see a comparison of an agent making charts for you based on a prompt and some data directly into Javasript it looks like, versus the same agent making the charts for you in Flint.
Re: Show HN: Microsoft releases Flint, a visualization language for AI agents
#146Earlier quoted context omitted.
When I first saw Claude generating PPT decks by writing Python code instead of making the XML directly, it was sort of an "aha moment" for me. This seems to be the path for many things. It also feels slightly limiting, and like a hack LONG term, but 100% correct approach for a while.
Do we have more details / documentation on this capability. Can we replicate the same using a generic agent skill with custom apps / harness. I think I saw some library named pptxjs? in the thinking traces of Claude. Or was it this: https://www.npmjs.com/package/pptxgenjs
Anthropic released their PPTX skill[0] a while ago, the relies a bit on python and pptxgenjs, depending on the usecase:
> Read/analyze content....python -m markitdown presentation.pptx > Edit or create from template....Read editing.md > Create from scratch...Read pptxgenjs.md
[0] https://github.com/anthropics/skills/blob/main/skills/pptx/S...
Re: Show HN: Microsoft releases Flint, a visualization language for AI agents
#147I like it.Impressive variations.
Re: Show HN: Microsoft releases Flint, a visualization language for AI agents
#148Earlier quoted context omitted.
Do we have more details / documentation on this capability. Can we replicate the same using a generic agent skill with custom apps / harness. I think I saw some library named pptxjs? in the thinking traces of Claude. Or was it this: https://www.npmjs.com/package/pptxgenjs
Anthropic released their PPTX skill[0] a while ago, the relies a bit on python and pptxgenjs, depending on the usecase: > Read/analyze content....python -m markitdown presentation.pptx > Edit or create from template....Read editing.md > Create from scratch...Read pptxgenjs.md [0] https://github.com/anthropics/skills/blob/main/skills/pptx/S...
Thanks -- will study and adopt this. Appreciate the pointer.
Re: Show HN: Microsoft releases Flint, a visualization language for AI agents
#149Earlier quoted context omitted.
What is it about json that models struggle with, in your experience? Is it syntax or structure?
It's often comes with missing keys, use wrong value type (e.g., list over dictionaries). Mostly a small model issue and open source models, they don't follow instructions on the structure guidance that well, and there is no easy way to do generation-time validation.
Hmm this is a problem that has long been solved with code completion in a JSON language server and a valid JSON schema.
By this I mean that you can basically tab-complete your way into a valid JSON document matching a particular schema.
Maybe there is a way to restrict the set of output tokens based on the current position in a JSON schema of a JSON document the model is predicting?