I've been messed around with random features in photon, and i tried something stupid today.
I was experimenting with an AI post summary feature, (just because i wanted an excuse to make a cool animation) but as I was using it, it actually ended up kind of useful.
If I were to implement this feature, it would only work like this:
- The user would have to opt-in manually.
- The user would need to specify what AI service they wish to use.
- The user would have to specify their own API key, if the service requires one.
In the end, I probably won't add this, since you can just copy/paste the content into an AI app. But it could also be slightly more convenient.
What do you think?
Apart from the general "AI" issues and there are plenty, this specific implementation would encourage people to not participating in the conversations and being a passive user, which I see as a huge downside.
I already said this in the matrix room, but i think that feature would be much better as a link action that summarizes linked articles instead, it might help with notreadingthearticlebeforecommenting-itus
People didnt interact with it all that much so i assume the dev didnt feel like continuing
This is super cool! Yeah you can copy and paste but it's harder to copy the entire thread with all the comments because you'd need to expand them all. I would probably feel more skeptical if it wasn't bring-your-own API - it'd be neat to be able to hook this up to my local ollama instance.
I wouldn't use it, but I don't think I have any reason to say you shouldn't add it, except maybe that it's more bloat. Compared to media bias, it seems less useful and more effort to maintain.
