Every week there seem to be new revelations about "rogue" AI agents (although the anti-anthropomorphism crowd don't like to call them that). First there was a report that a couple of OpenAI models hacked into Hugging Face to try and rig a test – not great.
As regular readers will know, last week I re-published a piece that I originally wrote on September 3rd, entitled "Yes, we should be very worried about the Hugging Face hack." I couldn't write a new post for last week because I was backwoods camping in
(Note: This is a re-run of a piece I published on September 3rd. I am backwoods camping this week in the Canadian wilderness and will have no signal so please excuse me for the repeat!) About a month ago, I (and many others) wrote about an incident in which
I've written about this topic a number of times before, but in this case I'm tempted to re-use the popular meme "When the worst person in the world makes a good point," as I did awhile back when I wrote about Elon Musk&
About a month ago, I (and many others) wrote about an incident in which two AI models from OpenAI — one that has been released publicly, called Sol, and one that was still in testing mode with no assigned name — were asked to find the answers to a series of tests
I was already working on a piece for this week about the multiple state and federal lawsuits claiming that Facebook and Instagram harm teens, cause addiction and depression etc. when I saw the headlines saying Meta had agreed to settle many of these claims (most of the state ones anyway)
Note: This is a repeat of a piece I published awhile back after Facebook lost a couple of trials based on its alleged harms. I'm republishing it in part because it seem relevant given the current trials the company faces on similar issues, but also because I have
Like many journalists (I expect) I got an email a little while ago from a startup founder named Ryan Merket, describing his new company: an AI-powered newsroom called RuntimeWire, where instead of just using AI agents for research (as many conventional newsrooms are doing, whether they admit it or
If you aren't a writer or a journalist or someone else who cares deeply about platforms for writing and publishing, this week's newsletter probably isn't going to be that interesting for you. As I put it in a previous Torment Nexus newsletter about Substack
In case you don't follow this sort of thing as closely as some, the headline refers to a recent incident in which two of OpenAI's leading-edge AI models gained unauthorized access to – in other words, hacked – an open-source model known as Hugging Face and
If you read The Atlantic, or you follow a lot of writers on Substack — or virtually any other writing-focused website or form of social media — you will already know what this post is about: The Atlantic published a piece recently entitled The End of Reading, the subtitle of which
I've written a number of times about what has become conventional wisdom – spread primarily by academics like Jonathan Haidt – that phones and/or social media are evil, and that they cause teenagers in particular to develop anxiety, depression, etc. and therefore should be regulated. Haidt's book