Home Speakers Scott Alexander Quote For 100 steps in a row, the AI can shoot delicate subtle ideas from layer to layer at the speed of light. Then there’s one step where it has to encode them into twenty-six glyphs invented by Phoenician turquoise miners in 1800 BC. Then it has to re-encode the Phoenician glyphs into delicate subtle lightspeed ideas before it can do anything else.
Words found on the source page on Nov 6, 2026.
More from Scott Alexander Blog The Specter Of Neuralese Sep 24, 2026
An AI with more real layers can actually be “smarter” in the conventional sense of the term. Blog The Specter Of Neuralese Sep 24, 2026
The science of reading these thoughts is a subfield of AI interpretability, which is still in its infancy. Blog Mysteries Of AI Generalization Sep 23, 2026
This story of misalignment says that LLM alignment makes AIs more aligned, RLVR makes them less aligned, and the exact level of alignment depends on how these two things interact or cancel out. But if this work generalizes, all the bad effects from RLVR get sequestered to RLVR like problems. Blog Mysteries Of AI Generalization Sep 23, 2026
This trains the AIs to be focused on task success, which naturally risks including things like reward-hacking, cheating, and single-minded pursuit of stated goals at the expense of ethical injunctions.
On the same topics Blog Why I still haven't bought into true RSI Sep 19, 2026
It is important to not confuse massive steps in inference-time scaling, a dynamic which should be fairly predictable, with being the outputs of RSI, which is highly uncertain. Podcast Dwarkesh Podcast: Agent swarms, alignment, and recursive self-improvement Sep 17, 2026
If you have five hours, you’re probably going to do a lot better. The AI models are pretty similar. Podcast Dwarkesh Podcast: Agent swarms, alignment, and recursive self-improvement Sep 17, 2026
So I was like, “I don’t think we’re going to get it in 2026, probably not in 2027, maybe in 2028.” So it did happen a lot faster than I expected. Podcast Dwarkesh Podcast: Agent swarms, alignment, and recursive self-improvement Sep 17, 2026
The concerning scenario is that, especially as these models are becoming more capable, we make them what we think is aligned, and they’re 99.9% aligned. Then we use these models to help us with the next generation of models, and they end up being 99.8% aligned. Then with each subsequent generation, we see an increasing degradation in alignment. Blog We Must Pace The Frontier Sep 14, 2026
If I was in charge of an alignment or interpretability team I would never run out of experiments to run and things to try, including having the time to properly talk to the models in the first place. Blog We Must Pace The Frontier Sep 14, 2026
Over the summer there was a step change, as Mythos and Astra started kicking off the early stages of recursive self-improvement (RSI). About this quote 
Quotesby Glyph