Meta will label AI content to help prevent deepfakes on Facebook and Instagram

[geekthegame]

https://www.axios.com/2024/04/05/meta-broader-ai-labeling

[beezlewax]

Sorry but how? They can't reliably tell the difference

[bamboozled → beezlewax]

I doubt they really mean it, generative AI content is a true boon for social media companies, an endless stream of "weird" quirky content for people to click on.

They mean they'll do their best.

[sandspar → bamboozled]

Young and savvy people will uninstall Instagram and TikTok if the apps get infested. Trend setting teenage girls are interested in watching humans talking, not in watching Beeple videos.

[sunaookami → sandspar]

I bet that most people (even tech-savy) can't reliably differentiate real content from AI content. Instagram and Threads for example are already full of AI content and AI "models".

[geekthegame → beezlewax]

not sure how, but they're going to need to. Otherwise AI content (which will improve) will take over their platform and ruin it

already happening on instagram, seeing more AI accounts pop up

[brevitea]

... one more statement we don't trust from Meta.