Thanks! This is a really important one. Only if we had a way to convert these pain points to action points and organize financial support for the effort to address them…

I must say that the masto5 post felt like really a big step in this direction, but the threadiverse has still to get the academic and institutional support mastodon has managed to achieve.

I wouldn’t think that your advertising of the podcast is so successful as to make people listen to it.

But somehow you seem to be leaking the idea that reasoning models are something different than autoregressive LLMs. Yes, they do have different fine-tuning and system prompts, but little beyond that. Subbarao Kambhampati has worked a lot on this, e.g. https://doi.org/10.1111/nyas.15339 or https://doi.org/10.1111/nyas.15125

I struggle to imagine what would be a possible meaningful answer.

“I’ve hand 15 posts and 6 comments in the 11 communities I moderate, but the mood on this part of the world is gloomy because of politics”?

Exactly, thanks for sharing the link. The interaction between threadiverse platforms and mastodon (sadly for pixelfed this does not work) is exactly as it is for fedigroups.social or what used to be a.gup.pe.

I don’t see this disputed so much as I see the need for “all science requires methodology”. I find it ridiculous that research in technical areas like LLMs has given up on any pretence of reproducibility.

That’s a very simplistic interpretation. A popular book that counters it is https://bookwyrm.social/book/29025/s/thinking-fast-and-slow

I’m not that convinced that not responding to “how’s everybody doing with their communities” is indicative of anything.

Would also be nice if they start explicitly treating the threadiverse as groups. This is already doable, but some usability sweeteners would make it much more accessible. It’s an absolute win-win for everyone on the fediverse.

That’s amazing progress. My next wish: being able to see past DMs with someone.

Consider https://surf.social/ as an option.