Welcome to Vuelta Abajo......& AI disclosure

[El Presidente]

AI ...what a connundrum. 

Videos such as this raise some important questions that can't be kicked down the road. 

At a minimum, shouldn't AI content (any), be clearly labelled as such to the viewer?

Is that too much to expect?  Otherwise in 5 years, how would you know anything you or your family are watching  is real? 

Platform screening for AI content and ensuring propper labelling couldn't be difficult. Fining content creators and platforms for non labelling of AI content (minimum standards to be determined) would be a first step. 

 

Kick it around 

[rckymtn22]

Bizarre video. World most expensive dog rockets. I wouldn’t spend $100 on those.

[Chibearsv]

9 hours ago, El Presidente said:

At a minimum, shouldn't AI content (any), be clearly labelled as such to the viewer?

I think all platforms should use AI to label all news and social posts. AI could grade each post on a scale of "Strictly for entertainment purposes only" or 1-5 with 1 being furthest from factual and 5 being closest to factual.

[NYGuido]

8 hours ago, Chibearsv said:

I think all platforms should use AI to label all news and social posts.  AI could grade each post on a scale of "Strictly for entertainment purposes only" or 1-5 with 1 being furthest from factual and 5 being closest to factual.

I am a lawyer specializing, in part, in AI work, including AI policy. This is way harder than you'd think because AI cannot, itself, verify factual accuracy unless it has been trained to respect certain sources over others. And, even then, it fails a lot of times when it goes out into the larger internet to gather second-level information for its answers. It absolutely is something folks are working on, but even that process has inherent bias based on what the developer chooses to designate as trustworthy. The best AI can do now is aggregate many sources and compare info and give you the most common answer as the most "accurate," which is better than nothing.

On the AI labeling front, this is something US lawmakers want to prioritize and--when talking solely about disclosing whether something used AI in its production--is fine under current 1st Amendment doctrine. Where it gets sticky (under a precedent called Zauderer) is requiring companies to disclose anything other than purely factual information about certain things. Requiring them to opine on stuff like accuracy might be seen as compelled corporate speech in a way that at least the 9th Circuit has called into question (see the cases against the CA AG by NetChoice and X).

[JPark3]

Giant cigar, with a thick ring gauge in the thumbnail, lines up with today's fad. Not entirely convinced it's AI. Will need to investigate further. 

[Lucas Buck]

I would LOVE it if all AI was labeled as such.