161–170 of 496 posts
It doesn't seem to do be able to do anything that Stable Diffusion cannot do, and I bet they put a ton more restrictions too. Other than OpenAI, most of these closed-source AI developments are quite underwhelming.
Earlier quoted context omitted.
And if you "learn" something and accidentally rewrite it verbatim? Thats what clean-room design is to protect against
Rewriting the code verbatim and distributing it would be a copyright infringement, yes, you do not have a write to distribute code written by other people That's completely different from reading and learning from code, which is what grondo described. Clean room design relies on this, in a clean room design you have one party read and describe the protected work, and another party implement it. That first party readi…
AI (e.g. copilot) has already been shown to break copyright of material in its training set. Thats the context of this whole thread.
Earlier quoted context omitted.
Still on the fence for whether or not you should be able to opt out of training (I'm sure many artists would love to "opt out" of humans looking at their art if the human intends to, or might, copy the artists' style at some point).
I can train a LORA on my own PC in less than an hour. Good luck opting out of that.
How long do we think this "big company gates access to an impressive AI model" moment will last? I wonder if generative models will become such a commodity that they cease to be a revenue driver or differentiator.
There will be moats around training data (Adobe probably has huge amounts of high-quality training data that isn’t available to the public), and around fine-tuning for specific fields. The more specialized the application, the less of a commodity it will be. And it remains to be seen how long it will take until there is an open-source model on the level of GPT-4. It may be harder than many expect, and the commercial…
Big AI's advantage is still being made. Every time a user hits "Feedback" they're fine-tuning this proprietary model. You can absolutely make an Open Source model with enough compute, but if it is only e.g. "95%" good relative to the paid one at "97%" or "98%" are you going to prioritize it? How many of you use Google because it is 2% better than Bing for example?
What this reinforces is that unlike with previous big innovations (cloud, iphone, etc), incumbents will not sit on their laurels with the AI wave. They are aggressively integrating it into their products which (1) provides a relatively cheap step function upgrade and (2) keeps the barrier high for startups to use AI as their wedge. I attribute the speed at which incumbents are integrating AI into their products to a…
The skeptic in me thinks it's more:
* The market is rewarding companies for doing X (integrating AI), so companies are doing X (integrating AI).
Song as old as time.
Earlier quoted context omitted.
I strongly suggest everyone to read this: https://techcrunch.com/2023/01/06/is-adobe-using-your-photos... I hope it's fair to say that they do train on your work.
We dont. More info here: >The insights obtained through content analysis will not be used to re-create your content or lead to identifying any personal information. https://helpx.adobe.com/manage-account/using/machine-learnin...
I hope you do continue doing so. I'm all but disappointed in others' approaches in this area, and it paints a very bad image for the potential of AI as tools.
Closed beta. Requesting access requires filling out a four page form. Product page is a bunch of hand-selected images. Nothing to see here.
https://www.youtube.com/watch?v=BkTXyY9cnEs
and there is a live stream later today here:
Earlier quoted context omitted.
Rewriting the code verbatim and distributing it would be a copyright infringement, yes, you do not have a write to distribute code written by other people That's completely different from reading and learning from code, which is what grondo described. Clean room design relies on this, in a clean room design you have one party read and describe the protected work, and another party implement it. That first party readi…
> That's completely different from reading and learning from code, which is what grondo described. AI (e.g. copilot) has already been shown to break copyright of material in its training set. Thats the context of this whole thread.
If an AI infringes on copyright then it infringes on copyright, that's unfortunate for the distributors of that code.
Humans accidentally infringe on copyright sometimes too. It's not a unique problem to machine learning. The potential to infringe on copyright has not made observing/learning/watching/reading copyright materials prohibited for humans, nor should it or (likely) will it become prohibited for machine learning algorithms.
How is this 'Firefly Model' trained and sourced? Will it be on the contents of the stock.adobe.com library? Clicking through the available pages it seems like a lot of 'coming soon' talk, so there's not really any detail about any of the underlying process.
I'm sure it's all of their content plus half of the web. The proprietary data they get from Behance, Lightroom, Photoshop and Illustrator (and soon figma) has to be a great advantage for them though.
https://helpx.adobe.com/manage-account/using/machine-learnin...
"The insights obtained through content analysis will not be used to re-create your content or lead to identifying any personal information."
Earlier quoted context omitted.
Maybe, maybe not. For instance, I'd like to make a game. But I don't have enough money to even hire an artist to help with concept art. So I don't get to the point where I can raise money off that art, and hire an artist to make game assets. Now I can generate all the concept art I want for free, and I can raise money off of that (wallets open faster with pretty pictures than with words). What am I going to do with i…
Except now you’re competing for the attention and disposable income of everybody else doing that. How are consumers going to tell your stuff apart from all the other AI placeholder games that will flood the market?
For someone like me, I can do a lot, but not everything. I've managed to get my own project to a stage where I had felt I would have to bring on more people to advance it much further. But I had lacked the funds to do so, and it's hard to get people to do things for free when they don't believe in it. It's also hard to get people to believe in something without seeing it. My project was very likely.
So that's where tools like stable diffusion and chatgpt come in. I'm now suddenly unstuck; I have a cheap tool to do work I wasn't capable of before, so I can now take the project further than I could have otherwise. Whereas before I might have abandoned it, now I can take it further and maybe get it to the point where I can hire people. The question now is: how many projects are now going to take off? Is there funding out there for them? Can they hire more artists than are displaced?