Shouldn’t be reading this before bed because it isn’t calming me down.

I don’t think I’ve ever read an interview as interesting as this one.

Really appreciate this essay. It puts things nicely and has the kind of personal investment that makes it relatable.

I found this very useful!

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

I wrote recently about how my concerns about (generative AI) are probably more about the broader Ellulian system of technique than the specifics of the technology. Here’s a passage from this article that makes a similar point better:

For some tasks, AI really is amazing; the tech behind things like machine-learning algorithms and large language models is ingenious, but the results always seem to be hawked the hardest by people and companies I don’t particularly like or trust. (Heck, Anthropic used one of my books to train its database, a sin for which it is now paying authors in court.) Give me the same sorts of tools but under my local control, governed by a Wikipedia-style nonprofit and trained on ethically sourced data, and I’d use them a lot more.

Amazing pictures—might have to replace my computer desktop.

So Doctorow already said this in the essay I linked to this morning, but it’s very striking how much this leak highlights Anthropic’s willingness to use our creative labor while simultaneously cracking down on any effort for others to do the same to them.

Mixed feelings about this. I think there are good reasons to be skeptical of ed tech at this level, but I also think that there is some unwarranted handwringing going on here. I have trouble untangling the two.

Hate this so much. The company statements are weaselly and gross. This also demonstrates one of the worst things about this whole issue. I would never buy Meta smart glasses, but I know two people in my life who own them, and the privacy burden of this product isn’t borne by the owner of the device so much as by those around them. Barf, barf, barf.

Facial recognition is always a scary technology, but it’s becoming even more so under the current administration.

Big Ellulian “image vs. word” vibes here.

A few thoughts:

First, it is almost comically mean to use the results of a project collecting AI tells to get LLMs to not sound like that. Like, of all the digital labor exploitations of AI, this might be the pettiest.

Second, AI detection is hard, and for all my concerns with AI, I think this is another good example of why policing its use can do more harm than good. I don’t blame the Wikipedia community for doing this project, but I would never recommend this approach in a classroom.

This is such a perfectly dumb oversight.

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

Bookmarking because this is an important point.

I genuinely don’t know what legal liability for generative AI products should look like, but arguing that the onus was on the kid and his family because of TOS strikes me as incredibly shitty, not to mention falling back on “look, we have a mission to benefit humanity by building AI, have you taken that into account?”

Very happy for Linkletter, but it’s shameful that Proctorio got away with as much nonsense as it did.

Finally got this book, so I guess I need to read it now.

This is a good article; it would be easy to just roll eyes at Cruz, but this goes further in explaining how Wikipedia actually works.

Just bookmarking everything I read on Sora for future grumpiness.

This is horrifying. Reading the headline is one thing, but reading some of the details is stomach-churning. I’m not a lawyer, and as disgusted as I am with this, I don’t know what legal liability should look like here. I feel more comfortable describing this as ethically bankrupt, though. I think I would have many fewer concerns about generative AI if it weren’t a platformized consumer product. Whatever the right legal response to this is, OpenAI has some moral responsibility for this sort of thing.

Again, I’m not sure copyright is the way to go in fighting immoral generative AI companies (that the ALA and EFF are on Anthropic’s side seems important to me), but “we have to be able to do this to be successful” still strikes me as such a hollow, self-serving argument.

Here’s Jacques Ellul on state funding of research:

The state demands that anything scientific enter into the line of “normal” development, not only for the stake of the public interest but also because of its will to power. We have previously noted that this will to power has found in technique an extraordinary means of expression. The state quickly comes to demand that technique keep its promises and be an effective servant of state power. Everything not of direct interest to this drive for power appears valueless.

This article is one more push in the direction of my finally subscribing to Kagi.