News for hackers. Stuff that matters.
Re:Adobe Firefly vs. MidJourney AI image generation by blibble on Monday April 03, 2023 @04:47PM (#35431045)
it's not learning anything

if you turn the temperature value down to zero it will produce the same thing every time given the same input (in the same way zlib or jpeg will)

it's a lossy compression function that is implemented using a neural net

copyright isn't removed if I take someone's existing image and re-save it as a jpg

Re:Adobe Firefly vs. MidJourney AI image generation by alxhill on Monday April 03, 2023 @05:17PM (#35431409)
This is false and represents a poor understanding of how these models work - they do abstract concepts and no you can't trivially get training images out.
Re:Adobe Firefly vs. MidJourney AI image generation by blibble on Monday April 03, 2023 @05:23PM (#35431484)
it's exactly the argument that the court cases against training on copyrighted works without permission are using

if the courts agree and this is ruled to be infringement then we'll left with products with level of quality we see here

meaning the technology will be an economic dead end

here's to hoping

Re:Adobe Firefly vs. MidJourney AI image generation by Robotbeat on Monday April 03, 2023 @09:13PM (#35433697)
Copyright is about copying. It is not about observing. Reading a copyrighted book isn't infringement. Writing out copies of it is, even if you don't use a computer or anything.
Re:Adobe Firefly vs. MidJourney AI image generation by DiscourseFan on Monday April 03, 2023 @11:13PM (#35434518)
>they do abstract concepts

I don't think, given that we don't even have a particularly full understanding of how human conceptual logic functions, that we can claim that even AIs are using it as well. It's only "abstracting" in the sense that it has labeled one million objects in its training set with the word "tree," and fuses many of those images together to form a general picture dependent on specific parameters made to limit its set (oak tree, winter tree, etc.)

But that is different from me or you using the word tree, which is just a signifier among signifiers, it stands for nothing but a negation of the very thing it points to in a certain set of symbolic relations. Humans communicate in the order of symbolic structures, our minds function much more like LLMs, creating multitudinous pattern relations. What you call "abstract concepts" are of a secondary order imposed to create rigorous exactitude overtop the riddled mess that we call the human psyche.

Re:Adobe Firefly vs. MidJourney AI image generation by RC_ITR on Monday April 03, 2023 @06:13PM (#35432049)
Ok, so let's assume it's using a Visual Transformer architecture, since AFAICT, we don't know.

What the model does is turn the picture into numbers in a clever way (embedding).

Then, the model compares the relationships between all the embedded tokens (self-attention).

Then the model does the fancy linear algebra stuff we've come to know and love (feed forward neural network).

What you're basically getting at with your temperature comment is "rather than having it semi-randomly pick one of the top n options that the feed-forward network recommends for which pixel goes where, we can always have the model simply choose the top answer each time."

That is not "we can make it reproduce training images"

The point of embedding is to see multiple examples of similar images and generalize from them. The process in and of itself breaks the ability to re-create a photo, unless the training set is very skewed.

If that's the argument of legal cases, they are likely to fail or win on ignorance.

Re:Adobe Firefly vs. MidJourney AI image generation by blibble on Monday April 03, 2023 @06:21PM (#35432126)
> That is not "we can make it reproduce training images"

didn't say that, all I said is it's a deterministic lossy compression function of the inputs (unless randomness is deliberately introduced)

and as such it should be treated the same way as zlib or jpeg for copyright purposes

Re:Adobe Firefly vs. MidJourney AI image generation by f_devd on Monday April 03, 2023 @06:36PM (#35432301)
> deterministic lossy compression function

This would only be true if it didn't mix or 'forget' any of the training data for each prompt, and only then for the exact same inputs of the training data. That is a lot of qualifiers that aren't practically applicable.

Slight side note: a deterministic lossy compression function should not always be treated the same as jpeg/zlib, as hash functions are technically also in that category.

Re:Adobe Firefly vs. MidJourney AI image generation by RC_ITR on Monday April 03, 2023 @06:35PM (#35432271)
I think you're really stretching the definition of deterministic here.

For example, switching training order of images has a known impact on outputs, but I don't see the equivalent in traditional compression.

I agree with you that it is computer code that runs on circuits (and is therefore deterministic in a very loose sense), but it is fundamentally incorrect to compare it to traditional image compression.

Re:Adobe Firefly vs. MidJourney AI image generation by pertymcpert on Monday April 03, 2023 @06:55PM (#35432505)
Those are two orthogonal concepts. The temperature has nothing to do with whether it's a lossy compression function.
Re:Adobe Firefly vs. MidJourney AI image generation by Robotbeat on Monday April 03, 2023 @09:09PM (#35433672)
Copyright doesn't have anything to do with determinism. It existed in a pure analogue world before computers and if you had implemented your neural net in a fundamentally non-deterministic way (such as an analog neuromorphic neural net of some sort), it wouldn't make any difference one way or another. A human artist with no computer whatsoever and only their brain and basic writing (or painting) implements can break copyright just fine by writing out (or painting) copyrighted works from memory, then distributing them. It's the PRODUCTION side that is the copyright infringement, not the observing side.
Re:Adobe Firefly vs. MidJourney AI image generation by avisser on Tuesday April 04, 2023 @08:18AM (#35438396)
I have no idea why your replies seem to think you're being outrageous. I've seen the image from one of the AI services that is a very clear almost-reproduction of a Getty Images photo.

I wouldn't want to be the defense's lawyer when that image gets shown to a jury.

https://www.theverge.com/2023/2/6/23587393/ai-art-copyright-...

Re:Adobe Firefly vs. MidJourney AI image generation by Robotbeat on Monday April 03, 2023 @08:49PM (#35433533)
The human brain is a lossy compression function implemented with a neural net.

Making absurd arguments is fun!

(The reason this is absurd is the number of parameters in the model is incredibly small compared to the size of the training set. You wouldn't even have enough bits per training image to keep a unique identifier for each training image, let alone a compressed representation. And to the degree there's redundancy in the training set, this is undesirable, actually counter-productive to the functioning of the model.)

(In a way, the human brain is actually more capable of copyright infringement since the number of parameters in the human brain is about 4-5 times that of a typical image generative AI neural net, and we're trained on a smaller and less generalizable data set. Humans can and do make copies of things from memory, which generative AI massively struggles with and can only do for corner cases.)