Hacker Neue

Preferences
CompoundEyes parent
I don't have access but it seems you can impose a friend into a video? Are we not rightsholders to our own likeness? It seems like a person should be able to block a video someone shares without their consent or earn revenue then if their likeness is used.
minimaxir
You have to explicitly opt into sharing your likeness with permission controls.

> person should be able to block a video someone shares without their consent

That is already implemented.

solid_fuel
> You have to explicitly opt into sharing your likeness with permission controls.

Ok... how is that supposed to work? I don't have an OpenAI account, there are no permission controls for me. Someone else could easily upload a picture of me, no?

pants2
No, you have to register yourself with a video where you're required to say a unique code.

So unless you've posted a video of yourself online saying every number from 1 to 99 they won't be able to copy your likeness

kouteiheika
This seems... pretty easy to get around? There are already open weight models which can take any photo and audio and make a video out of it with the character speaking/singing/whatever, and it runs on normal consumer hardware.
pants2
So you wouldn't know what the three numbers are ahead of time, you'd have to be using a real time face replacement model (or I guess live-switching between pre-rendered clips) and somehow convince the app that you're the iPhone selfie cam.

But at that point you might as well just use WAN 2.2 Animate and forget about Sora.

solid_fuel
That's more than I expected from them, genuinely. But it still doesn't seem like a very solid solution. I wonder how much variation in look and voice it accepts?

My partner likes to cosplay, and some of the costumes are quite extensive. If they want to generate a video in a specific outfit will they need to record a new source video? The problem exists in the other direction, too. If someone looks a lot like Harrison Ford, will they be able to create videos with their own likeness?

I wonder how this extends to videos with multiple people, as well. E.g. if both my friend and I want to be in a video together.

WmWsjA6B29B4nfk
It’s not like making a video of someone saying a number, given a single photo and any voice sample is a very difficult problem today. We can just fast-forward a few weeks into a world where this „registration“ is already broken.
noduerme
So only the heads of companies who lead shareholder meetings are vulnerable to this exploit? Cool.

This item has no comments currently.

Keyboard Shortcuts

Story Lists

j
Next story
k
Previous story
Shift+j
Last story
Shift+k
First story
o Enter
Go to story URL
c
Go to comments
u
Go to author

Navigation

Shift+t
Go to top stories
Shift+n
Go to new stories
Shift+b
Go to best stories
Shift+a
Go to Ask HN
Shift+s
Go to Show HN

Miscellaneous

?
Show this modal