No spam, ever. We'll never share your email address and you can opt out at any time. Already a subscriber? Log in
As usual, @tressiemcphd.bsky.social is spot on with her analysis of our political moment. Here, asking why Democrats aren't capitalizing on rage against data centers. I was happy to chip in my own two cents on why Bernie is missing the mark. www.nytimes.com/2026/06/12/o...
Grateful to The Verge for publishing my essay on why large-language models are not going to achieve general intelligence nor push the scientific frontier. www.theverge.com/ai-artificia...
I had given Anthropic a lot of credit for turning down the DoD and its trillions of dollars, especially as it seemed to be the only example of any AI company making any financial sacrifices for moral principles. Of course, it turns out to be not really true. www.axios.com/2026…
How many times do I have to tell you: "These AI models … will never say 'no'" ought to be regarded as a Failure state, not a desired performance metric. www.wfae.org/united-state...
www.educationnext.org/generative-a...
Fortunately, House Democrats were able to block the bill to reauthorize FISA. But this fight to protect our civil liberties isn’t over. Email your Members of Congress and demand they don’t reauthorize FISA without major reforms like closing the loophole that allows warrantless…
www.theatlantic.com/ideas/2026/0...
"What a time to be alive! We are finally freeing you from the rat race and placing you on the rat unemployment line. Is your dream to be an artist? Well, it is the dream of your employers to push a button that extrudes art. And guess who has more funding for their dream?"
Don't worry you can fix all problems with language modeling with more data and harder RL
For all the money we spent doing AI we could afford to hire and pay more mathematicians
Not shocked at all that the guy who is asking chatgpt advice for building a stupid app is also in the wrong seat from his assigned seat
I am very excited to wind down my language modeling career because (1) it isn't fun anymore and (2) the AI guys are too apologetic about slop anyway
I am actually amazed at what the notes my students took into the final exam look like. A lot of AI slop. Most references aren't in my slides or the textbook
Damn you're telling me the one person I know who went to Harvard Law works for Anthropic now? Damn
AI has exacerbated the feeling that you can get far by bullshitting but like. You shouldn't be able to do that in college lol
It's really cool that the NSF spent some of its last money in SBE on pretty emptyheaded stuff because "AI"
Man I hope Anthropic drops off the face of the earth
it's very cool that everyone's favorite implementation for extracting surprisals is so slow it takes longer to compute probabilities with it than it does to fine-tune a model to criterion on the same dataset
Center = CJ. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.
