/tech/ - Technology

haven't watched that video but from reading the wiki it seems like this is just the digital equivalent of having a serial number on a gun. The point isnt to trace anonymous shitposts its to basically act like the old twitter verification badge but for information, or like an MP5 checksum. Anyone who wants to make a meme can just strip the metadata like you're already supposed to do with JPGs.

Just having a serial number on a gun doesn't prevent people from filing off that serial number illegally. For this it mainly seems against impersonation, so you can make whatever you want but Tom Cruise can sue the deepfake guy if he tries to claim that that's the real tom cruise and he actually said that.

This is a nothingburger even if passed. However its still a bad idea since it probably can't be enforced in practice.

>>20956
It seems like its only function is to track the provenance of information porky considers problematic with the pretext based on the disinformation hoax hysterics the MSM peddles.
Just looking at who's established this makes it clear this is part of an attempt to control the internet but through private sector standardization so they just claim "it's a private company, they can do what they want :)" when it's inevitably used against dissident voices.

Brought to you by Adobe, BBC, Intel, Microsoft, Sony to keep "online misinformation" (AI-generated content) in check with what amounts to DRM for digital media. It's pretty easy to get spooked by all of this.
Now to be fair, it's an open standard, so maybe the comparison to DRM isn't entirely accurate. They also claim it's opt-in, but I doubt it will stay that way. As the video >>20955 mentioned, Steam is already cracking down on games with AI-generated content so requiring C2PA seems like a logical next step for them, as with other web services.

Here's the technical specification by the way (perhaps someone can upload it): https://c2pa.org/specifications/specifications/1.3/specs/_attachments/C2PA_Specification.pdf
Here are a few key parts from the introduction (chapter 1):
>C2PA specifications SHOULD NOT provide value judgments about whether a given set of provenance data is 'good' or 'bad,' merely whether the assertions included within can be validated as associated with the underlying asset, correctly formed, and free from tampering.
>The identity of a signatory is not necessarily a human actor, and the identity presented may be a pseudonym, completely anonymous, or pertain to a service or trusted hardware device with its own identity, including an application running inside such a service or trusted hardware.
>The creators and publishers of the media assets always have control over whether provenance data is included as well as what specific pieces of data are included.
Another thing to take note of is chapter 18 which lays out the team's assessment of the "Threats and Security Considerations" and "Harms, Misuse, and Abuse". They basically say they will address these concerns sometime in the future. Pretty unsatisfactory in my opinion.

Here's a super simple explanation by CAI (a partner organization) for those of you who need it: https://contentauthenticity.org/how-it-works
Here's a guide on how not to do a case study (pro-tip: do not go off on irrelevant tangents) https://contentauthenticity.org/case-study

>>20959
>When you see the CAI info icon (the small i in the circle), you can inspect the image and dig into its edit history. If you click on the info icon on an image on this webpage, you’ll be able to see who took a photo, where, when, and with which device. You can see the tools they’ve used in Photoshop to edit the image and, by clicking the “View more” button, you’ll even be able to compare versions. This adds a layer of transparency and veracity to images that has never before been possible, so that any viewer of content on a news article or social media post can evaluate what they’re seeing and decide whether or not to trust it.

The problem is what happens if there was no edits made in the software that makes this meta data as all edits were made prior and imported in. For example what is Adobe going to do when you import an analog video signal in regards to if it is authentic? It seems like the whole endeavour has multiple ways to circumvent.

there's already a bajillion metadata standards, this does not sound very different
>le social credit
can this meme die already?
he's woke on the mars conspiracy though, I'll give him that

>>20971
The social credit meme is legitimate to the degree that it's the west seeking to implement such a system instead of China.