@zbyte64 from what I understand, you’re referring to the process at scale—the amount of information the AI can take in is inhuman—which I’m not disagreeing with
None of which is relevant to my original point: the scale of their operations, which has already been used countless times in copyright law
The scale at which they operate and their intention to profit is the basis for their infringement, how they’re doing it would be largely irrelevant in a copyright case, is my point
I don’t understand how when I say “agency” or “an aspect of the process” one would think I’m talking about the volume of information and not the quality.
@zbyte64 1) In no way is quality a part of that equation and 2) In what other contexts is quality ever a part of the equation? I mean I can go look at some Monets and paint some shitty water lillies, is that somehow problematic?
@zbyte64 data quality, again, was out of the scope of what I was talking about originally
Which, again, was that legal precedent would suggest that the *how* is largely irrelevant in copyright cases, they’re mostly focused on *why* and the *scale of the operation*
I’m not getting sued for copyright infringement by the NYT because I used inspect element to delete content to read behind their paywall, OpenAI is
@zbyte64 where am I wrong? The process is effectively the same: you get a set of training data (a textbook) and a set of validation data (a test) and voila, I’m trained
To learn how to draw an image of a thing, you look at the thing a lot (training data) and try sketching it out (validation data) until it’s right
How the data is acquired is irrelevant, I can pirate the textbook or trespass to find a particular flower, that doesn’t mean I’m learning differently than someone who paid for it
@zbyte64 I can get bad data through a pirated out of date textbook and that can affect my output, but the process of actually learning the information didn’t change
Do we assume everything read in a textbook is correct? When we get feedback on drawing, do we accept the feedback as always correct and applicable? We filter and groom data for the AI so it doesn’t need to learn these things.
@zbyte64 do I somehow owe Cheverny money because I did that? Would someone who did it better not owe them money? Where does quality come into play when determining infringement?
@zbyte64 I still have yet to understand how your original comment fits into the original conversation of infringement, now I’m really just along for the ride
@zbyte64 from what I understand, you’re referring to the process at scale—the amount of information the AI can take in is inhuman—which I’m not disagreeing with
None of which is relevant to my original point: the scale of their operations, which has already been used countless times in copyright law
The scale at which they operate and their intention to profit is the basis for their infringement, how they’re doing it would be largely irrelevant in a copyright case, is my point
I don’t understand how when I say “agency” or “an aspect of the process” one would think I’m talking about the volume of information and not the quality.
@zbyte64 1) In no way is quality a part of that equation and 2) In what other contexts is quality ever a part of the equation? I mean I can go look at some Monets and paint some shitty water lillies, is that somehow problematic?
If we’re using your paintings as training data for a Monet copy, then it could be.
Are we even talking about AI if we’re saying data quality doesn’t matter?
@zbyte64 data quality, again, was out of the scope of what I was talking about originally
Which, again, was that legal precedent would suggest that the *how* is largely irrelevant in copyright cases, they’re mostly focused on *why* and the *scale of the operation*
I’m not getting sued for copyright infringement by the NYT because I used inspect element to delete content to read behind their paywall, OpenAI is
I was narrowly taking issue with the comparison to how humans learn, I really don’t care about copyrights.
@zbyte64 where am I wrong? The process is effectively the same: you get a set of training data (a textbook) and a set of validation data (a test) and voila, I’m trained
To learn how to draw an image of a thing, you look at the thing a lot (training data) and try sketching it out (validation data) until it’s right
How the data is acquired is irrelevant, I can pirate the textbook or trespass to find a particular flower, that doesn’t mean I’m learning differently than someone who paid for it
@zbyte64 I can get bad data through a pirated out of date textbook and that can affect my output, but the process of actually learning the information didn’t change
Do we assume everything read in a textbook is correct? When we get feedback on drawing, do we accept the feedback as always correct and applicable? We filter and groom data for the AI so it doesn’t need to learn these things.
@zbyte64 do I somehow owe Cheverny money because I did that? Would someone who did it better not owe them money? Where does quality come into play when determining infringement?
@zbyte64 I still have yet to understand how your original comment fits into the original conversation of infringement, now I’m really just along for the ride