Time Nick Message 14:41 sfan5 merging #17376, #17352 in 15m or more 14:41 ShadowBot https://github.com/luanti-org/luanti/issues/17376 -- Use mesh cache when building node visuals by sfan5 14:41 ShadowBot https://github.com/luanti-org/luanti/issues/17352 -- Proposal: Luanti generative AI policy by v-rob 16:33 sfan5 done 16:35 MTDiscord In earth time as well 👍 19:07 MTDiscord Anthropic is giving away Max plans for OSS devs: https://www.anthropic.com/claude-for-oss-terms 19:08 MTDiscord I applied and got accepted, which means that I now have an absurd amount of tokens at my disposal over the next 6 months 19:08 sfan5 cool, free slop 19:09 sfan5 in all seriousness: maybe some sort of automated code review would be useful? 19:09 sfan5 and the documentation we'd have to write to make AI understand what's important would even be useful in isolation 19:10 MTDiscord indeed, i have some documents locally, but i never get around to polishing them so i can PR them to docs.luanti.org 🙃 19:11 MTDiscord luatic: I thought you was a megabrain developer for whom LLMs are simply unneded and useless… 19:12 MTDiscord Well-well it's all just a tool which can be necessarily useful if handled properly 19:12 MTDiscord i like to use my brain, but i can't discount the usefulness of LLMs. a bug found by a LLM is still a bug. 19:13 MTDiscord the current capabilities of LLMs in some aspects (e.g. vulnerability scanning) are pretty crazy 19:14 MTDiscord sadly Fable has been limited even more since I last ran it so I don't think I'll manage to coax it into another run. 20:14 MTDiscord hmm, do you think they'd let me qualify on the basis of being an Apache committer ("listed committer of a recognized foundation") and then use it to contribute to Luanti instead? My company pays for me to contribute to Apache, but I don't have a model to use for other personal projects. 22:09 MTDiscord I haven't read the terms too closely, but from my understanding if you get accepted, it is not tied to any specific project.