Do you think that Lemmy should incorporate generative AI features? If so, what would you like to see?
Do you think that Lemmy should incorporate generative AI features? If so, what would you like to see?
This thought entered my mind today as I came across a thread on Quora, and noticed that they have added a feature where ChatGPT would have a go at answering the question.
Today alone I have used a few varying "AI" tools, including one which automatically paraphrases text for you, one which analyses your writing in SwiftKey as you type, and of course the big players like Bard and Bing Chat. It got me thinking about whether these features are actually valuable, and if we would start to see them on this platform.
You're viewing a single thread.
Wh... Why? What do you imagine it adding to a social space?
What do you think it adds to a keyboard (swiftkey)? The feature is there regardless whether we like it or not. The question is whether you think such a thing should be on the site, and in what form, not whether one user considers to be useful or not.
One thing that it could do is analyse your perceived tone, and suggest edits to come across less aggressive, or more inquisitive for example. Such a thing could well make certain social places a more pleasant forum to use.
Why should that be part of lemmy? If you really wanted such feature it should be independent something like plugin into your browser. Same way as you would use plugins for things like spell checking.
I didn't necessarily mean hard baked into ActivityPub, Lemmy or any particular instance. It could take many forms, some more prevalent and noticeable than others. In a lot of use cases, AI/ML software is quietly ticking away behind the scenes, and you would never know.
What sort of independent plugins do you think we might see? Would you welcome such options, or prefer to reject them?
One thing that it has added to keyboards is making it much more difficult to use purposefully creative spelling or grammar.
For every problem they fix, they restrain creativity, because they themselves are constrained to their training set.
The Internet already has a problem with tone policing. Maybe we don't need mathematical models burning the planet to tell me that in pissed off.
If people want a moderated space, they can request some moderators.
Also, you didn't answer my question. Examples of what cna be done are not the same as answering why a thing should be done.