You know what the difference is, trying to act otherwise is just being obtuse.
How is this different from a human doing an impersonation?
Laticauda@lemmy.ca 1 year ago
TheGrandNagus@lemmy.world 1 year ago
Can you seriously not answer that question yourself?
tillimarleen@feddit.de 1 year ago
well, you seem to have trouble doing it
stopthatgirl7@kbin.social 1 year ago
You could say it’s not, which means in US law at least, it’s settled and they could be sued.
echodot@feddit.uk 1 year ago
There was a difference between complete duplication and impersonation for the purposes of satire.
Fordiman@programming.dev 1 year ago
Largely? The lack of convincing emotional range.
RizzRustbolt@lemmy.world 1 year ago
Can’t fake timbre.
photonic_sorcerer@lemmy.dbzer0.com 1 year ago
Because it can be done fast, reliably and at scale.
ArmokGoB@lemmy.dbzer0.com 1 year ago
Our entire society would collapse if we couldn’t do things fast, reliably, and at scale.
idiomaddict@feddit.de 1 year ago
Yes, but if “things” is replaced by scamming artists, that’s a shitty soxiety
ArmokGoB@lemmy.dbzer0.com 1 year ago
Artists aren’t being scammed. They’re being replaced by automated systems. It’s the same thing that happened to weavers and glassblowers. The issue isn’t that their job is being automated. It’s that people replaced by automation aren’t compensated. Blame the game, not the players.
BraveSirZaphod@kbin.social 1 year ago
I don't think it's a particularly mentally challenging concept to understand that we're not upset about the general concept of doing things at scale, and that it depends on what the thing in question is.
For instance, you'd probably not be terribly upset about me randomly approaching you on the street once - mildly annoyed at most. You'd probably be much more upset if I followed you around 24/7 every time you entered a public space and kept badgering you.
photonic_sorcerer@lemmy.dbzer0.com 1 year ago
Yes, but this is a new tool with new implications.
cloudy1999@sh.itjust.works 1 year ago
This, and it’s not a human. All these analogies trying to liken a learning algorithm to a learning human are not correct. An LLM is not a human.