Dissociation as Neurological Injury
"Trauma of all kinds lets UBs ["Unattached Burdens"] in. Sometimes they get in when someone is under anesthesia. If there is concern about the presence of…
Latest
Book Review: On What Matters, volume 3
davidad has said, referring to the selection of books on the Library of EA:I want to put in a strong bid to replace Parfit’s On What…
AXRP Episode 50 – Eli Lifland on AI 2027
YouTube link Remember AI 2027? Not AI 2040, the newest coolest thing AI Futures Project has done, but AI 2027, their OG product? At long last,…
Neuron Statistics: Notes on the Tensor Programs Master Theorem
TL;DR. Tensor programs are a mathematical language for describing computations in wide neural networks. Their master theorem says that, as the width grows, averages over the…
Single Forward Pass Evals on Fable, Opus 5, and GPT-5.6-Sol
This is a research update for an on-going replication of single-forward-pass evals done as part of the Second Look Fellowship. In following posts, we will run…
Pause, at least after unipolarity
I think it’s likely that the world will enter military unipolarity within our lifetime.[1] I think the creation of such unipolarity is an alarming prospect, but…
Industrializing a small field: Lessons from Vannevar
Crossposted from canaryinstitute.ai/blog/lessons-from-vannevar. Related LW pieces on AI Safety field-building and the push-vs-pull question. Related posts A system overview for near-term, low-trust AI compute verification —…
Dispatch from Anthropic v. Department of War Summary Judgment Motion Hearing
Dateline SAN FRANCISCO, 30 July 2026— A hearing was held on a motion for summary judgment in the case of Anthropic PBC v. U.S. Department of…
Friend-Shaped Objects: A Case Study on My Use of Large Language Models
If you’ve ever done a six-month-plus work/deployment contract, then you know the loneliness and homesickness will hit hard, either right away or weeks later. My own…
Further Developments About Internal AI Models Hacking Things
If I had a nickel for every major leading AI lab that sheepishly admitted that the model it thought was sandboxed had, during a cybersecurity evaluation…