Scott Alexander of AI Blog Astral Codex Ten: Why OpenAI Astra’s Layer Recurrence Alarms the Safety Community—Technical Equivalence Does Not Equal Normative Safety
OpenAI Chief Scientist Jakub Pachocki argues that the Astra model gains only twice the computational depth of GPT-4 through its recurrent layers, with no alignment difference compared to building a larger Transformer directly. In a September 2026 article, Scott Alexander explains why this technical defense may be insufficient from a governance perspective: if the safety community lacks clear taboo boundaries regarding layer recurrence, each breach justified by “just a few more layers” erodes the taboo itself.
OpenAI released the Astra model in September 2026, adopting a new architecture that loops its own internal Transformer layers rather than writing intermediate reasoning results as English ‘chain-of-thought’ notes as it did in the past.
To understand this debate, one must first distinguish between three types of reasoning. The first is a standard Transformer with a limited number of layers, which must write out its intermediate reasoning as readable chain-of-thought text. The second is a model like Astra that uses recurrence to simulate more layers, performing more computation within a single forward pass without outputting chain-of-thought. The third is ‘true neuralese,’ which uses no chain-of-thought at all; the AI can think secretly for an arbitrary duration, undetectable by humans.
The defense offered by OpenAI Chief Scientist Jakub Pachocki is that Astra’s computation graph depth is only about twice that of GPT-4, meaning there is no alignment difference from a technical standpoint compared to directly building an 180-layer Transformer. Over the past few years, companies have been adding layers: GPT-2 had 48 layers, GPT-3 had 96, and GPT-4 had approximately 120; no one claimed these were dangerous. Astra simply achieves similar depth through recurrence.
Alexander argues that this technical argument may be correct on its merits: recurrent layers are indeed less effective than truly new layers because recurrence merely reuses the same knowledge repeatedly, whereas new layers provide additional storage capacity to accommodate more knowledge and modes of thought.