TL;DR
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
Astral Codex Ten published its November 2026 links roundup, a collection spanning AI and disaster forecasts, essays, online debates and cultural observations. The author says the links have not all been independently verified and that readers often find errors in posts of this kind.
Astral Codex Ten has published its November 2026 links roundup, a collection of commentary and outside material touching on AI risk, California flood forecasts, the future of work and online debates. The author cautions that the links have not all been independently checked and that commenters typically identify errors or misleading items in each monthly post.
The roundup includes a forecast attributed to Sentinel, an organization described by the author as forecasting global disasters. Sentinel predicts a 25% chance that a rogue AI will exfiltrate its weights over the next 12 months. This is a forecast, not a report that such an event has happened. The post does not provide the forecast’s full methodology or define the scenario in enough detail to independently assess it.
A separate item relays a warning from a Sentinel-affiliated forecaster that California is expected to have a wetter-than-normal winter. The source refers to a 2.5% chance of a major flood, then presents interpretations from an AI forecasting system: an 8% chance of a flood causing at least $10,000 in damage to at least 100,000 homes, and a 2% chance of that level of damage across at least one million homes. The author also reports a 5% forecast for an ARkStorm this year, compared with a stated 1% baseline. These figures are presented in the post, not as independently verified official forecasts.
Other selections include an essay by Clara on how people might spend time in a world without work, commentary on the so-called Wow! signal, a discussion of Soviet acronym-based names and a reported NVIDIA agreement to buy Hugging Face for $12.9303 billion. The author says the purchase price contains a numerical reference to the hugging-face emoji, and estimates that the joke may have shifted the price by about $70 million if the alternative would have been an even $13 billion. That estimate depends on the author’s hypothetical comparison.
Forecasts Span AI and Flood Risk
The roundup brings together forecasts with possible consequences for AI security and disaster preparation, but readers should distinguish the claims from established events. A 25% forecast of AI weight exfiltration does not show that a system has been compromised, and the flood probabilities do not establish that a damaging flood will occur. They are estimates whose value depends on definitions, forecasting methods and how they are updated over time.
The post’s practical point on California is that residents may want to check roofs, gutters and other preparations ahead of a wetter-than-usual winter. That is the author’s recommendation, grounded in the forecast as described. The wider significance is that low-probability scenarios can be difficult to communicate: the author argues that a 5% chance of a very large disaster merits attention even when it is not likely enough to dominate public discussion.
The other links show the roundup’s broader role as a curated reading list rather than a single developing news story. Its selections connect current forecasts and technology business news with arguments about social status, work and the communities that produce online commentary.
Popular right now
Breville BES870XL Barista Express Espresso Machine, Brushed Stainless Steel
Dyson Cyclone V10 Origin Cordless Vacuum
KitchenAid 5.5-Qt Bowl-Lift Stand Mixer, Porcelain White (KSM55SXXX)
As an affiliate, we earn on qualifying purchases.
How the Monthly Links Work
The post is part of Astral Codex Ten’s recurring links format: the author collects material from other writers, social-media accounts and discussion threads, then adds reactions or summaries. The November edition ranges from short curiosities, such as an inhabited island reportedly covered by one house, to longer arguments about life without paid work and the culture of the Berkeley rationalist and effective-altruist communities.
That format matters when reading the entries. Some items are descriptions of linked articles; others are the author’s own judgments, estimates or jokes. For example, the proposed implications of a future in which AI performs all productive work are commentary on an essay, not a settled prediction. Likewise, the roundup’s account of a possible NVIDIA-Hugging Face purchase and its emoji-related price explanation are reported claims in the source material, not independently documented transaction details here.
The author specifically warns that readers commonly find two or three incorrect or misleading links in a links post. The author says corrections are made when noticed but cannot guarantee every issue has been caught by publication time.
“I haven’t independently verified each link.”
— Astral Codex Ten author
Forecasts and Link Details Need Checking
The source material does not include independent verification for the linked items, and the author explicitly says the roundup may contain wrong or misleading links. It is not clear from the post how Sentinel defines rogue-AI weight exfiltration, what evidence supports its 25% estimate, or how the forecast will be evaluated after the 12-month period.
The California figures also require care. The post relays a wetter-winter prediction and several probabilities, including estimates generated by an AI forecasting system, but does not provide enough detail to establish that the different flood scenarios use the same definitions or data. It is also unclear whether the NVIDIA-Hugging Face agreement described in the roundup has been formally completed, or whether the quoted purchase figure and the author’s estimate of the emoji joke’s effect have been independently confirmed. The source offers no transaction documentation or detailed terms.
Watch for Corrections and Updates
The author says corrections will be made as errors are noticed and that important corrections may be highlighted later. Readers can check the original post for updates and should treat its forecast figures as attributed estimates rather than verified outcomes. The stated AI forecast covers the next 12 months, while the California predictions concern the coming winter; the source does not give a specific update schedule for either.
Further confirmation would require checking the linked forecasting pages, relevant California weather and emergency-management information, and any formal announcement or filing concerning the reported NVIDIA-Hugging Face deal. Until then, the roundup is best read as a set of leads and commentary, with the author’s own warning about verification kept in view.
Key Questions
What is the November 2026 links post?
It is Astral Codex Ten’s monthly collection of links and commentary on subjects including AI, disaster forecasts, science, work and online communities.
Did Sentinel report that an AI has stolen its weights?
No. The roundup attributes a 25% chance forecast over the next 12 months to Sentinel. It does not report that weight exfiltration has occurred.
What does the post say about California flooding?
It relays a prediction of a wetter-than-normal winter and several flood-risk estimates, including figures attributed to an AI forecasting system. These are forecasts described by the author, not confirmation that a flood will happen.
Are all the links verified?
No. The author says the links have not all been independently verified and that commenters typically find around two or three wrong or misleading items per links post. The author says corrections are made when identified.
Source: astralcodexten.com
