Stories like this should take a prominent place in every discussion about generative AI.

I will (almost) always post articles that make me angry about NCII.

Helpful thinking and important suggestion.

To echo a comment I made earlier today, I will always hold tech companies morally responsible for the harms they cause, but I get a lot less sure about legal responses. Do we trust this administration to handle NCII properly?

Audrey Watters once compellingly argued that metal detectors are edtech. I think we now have a responsibility to treat AI nudifier apps as edtech, too.

So gross. I don’t think we can talk about generative AI without talking about this.

Look, I’m open to the argument that there are legitimate, good uses of generative AI, but I think anyone making that argument needs to address stuff like this.

Holy crap, these numbers. This passage really stood out:

“This is industrial-scale abuse of women and girls,” said Imran Ahmed, the chief executive of the Center for Countering Digital Hate, which conducts research on online hate and disinformation. “There have been nudifying tools, but they have never had the distribution, ease of use or the integration into a large platform that Elon Musk did with Grok.”

Good overview of an enraging story.

Depressing read with interesting details about why Grok is bad at this.

Oof, this line:

what is clear to me from following this Telegram community for a couple of years now is that nonconsensual sexual images of real people, including minors, is the cost of doing business with AI image generators

Bookmarking all these articles on Grok for rage fuel.

What a disaster.

Are we willing to pay this price in order to have some neat image generation tools? (I’m not.)

Look, nothing really new in here (Clearview should have made parents rethink the same ages ago, etc.), but yes, AI should get parents to be a hell of a lot more careful with posting pictures of kids.

I hadn’t thought about these two trends (cracking down on adult content, and Grok being Grok) being in tension with each other, and I appreciate what this article does to make that clear.

Unsurprising but disappointing.

There are, of course, social benefits to open sourcing powerful tools like these ones. However, I’m reminded of “open source” Android, which is a deliberate business decision that benefits Google—and of how many NCII-generating tools are based on open weight/open source models. gift link

I’m glad that someone is doing this white hat work, but I hate that we live in a world where someone has to.

Quel monde pourri qui attend nos enfants.

Gonna keep posting (almost) every article I read on NCII and generative AI.

This line (from a study quoted in the article) stood out:

The open-source nature of TTI technologies, proclaimed as a democratizing force in generative AI, has also enabled the propagation of models that perpetuate hypersexualized imagery and nonconsensual deepfakes.

Open sourcing generative AI solves some problems but creates others.

On, look, it’s two of my least favorite things about generative AI (NCII and raking in money without concern for ethics) IN THE SAME STORY.

Surely this is a reasonable price to pay for the Nazi-praising Grok to “discover new physics” within the next year, as Elon promised last night.

This kind of thing is why I hate “the genie is out of the bottle” arguments. I can’t help but hear them as “yes, people are going to create more CSAM, but all we can do is instead teach people to use these tools more responsibly.”

This problem makes me so angry, and while I appreciate this article’s exploration of different policy solutions, they also feel overwhelming to me because so many of them come with problems of their own.