Adobe Firefly Video Model

blog.adobe.com

6 points·leet··4 comments

Will temporal stability ever be a solved problem? A lot of these shots are still pretty "goopy" and present more action than really exists or just shuffles around constantly. The dandelion and reindeer examples are jarringly inorganic and would probably get identified as AI by most people who see it, I'd imagine.

> our community has generated over 12B images

They say it came out March 2023, so 1M images per hour.

Very active, this community.

Given that they can train their models on your Photoshop, Illustrator, After Effects and Premiere projects according to their ToC, this might not even be incorrect.

> Given that they can train their models on your Photoshop, Illustrator, After Effects and Premiere projects according to their ToC,

No, that is not correct.

https://www.adobe.com/ai/overview/firefly/gen-ai-commitments...

> Adobe Firefly is trained on a dataset of licensed content, such as Adobe Stock, and public domain content where copyright has expired. Adobe Stock content is covered under a separate license agreement, and Adobe compensates contributors for the use of that content.

> We do not train and have never trained Adobe Firefly on user content.

> We recently updated our General Terms of Use to explicitly spell out this commitment (Sections 2.2F and 4.3C2).

From the TOU:

https://www.adobe.com/legal/terms.html

> (F) Generative AI. We will not use your Local or Cloud Content to train generative AI models except for Content you choose to submit to the Adobe Stock marketplace, and this use is governed by the separate Adobe Stock Contributor Agreement.

> Generative AI. We will not use your Content to train generative AI models except for Content you chose to submit to the Adobe Stock marketplace, and this use is governed by the separate Adobe Stock Contributor Agreement.

(I work for Adobe)

Reply on news.ycombinator.com