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 a varied October 2026 links roundup, covering AI-generated art and translation, a bot’s reported win in the Metaculus Cup, and commentary on marriage, suburbs and polyamory. The author cautions that individual links have not all been independently verified and may be corrected.
Astral Codex Ten has published its October 2026 links roundup, highlighting a bot’s reported first-place finish in the Metaculus Cup alongside items on AI art, literary translation, marriage and urban life. The post is a collection of links and commentary, not a single investigation, and its author warns that some items may be wrong or misleading.
The roundup says the forecasting bot Laertes took first place in the September Metaculus Cup. Citing an account published by The Economist, the author reports that Laertes was developed by Jeffrey Liang, who estimated that he spent less than 150 hours and a few thousand dollars on computing power and data. The post also names Mantic, FutureSearch and Preseen among other high-performing bots, and says ACX reader Ben Shindel placed first among human participants.
The AI items also include examples of generated or AI-assisted creative work. The author praises an AI image inspired by Hieronymus Bosch, while acknowledging that its reliance on a human artist’s work complicates whether it should be treated as AI art alone. Another item describes an attempt to translate the Chinese novel Romance of the Three Kingdoms into rhyming pentameter, as well as an English rendering of the opening 100 lines of the Odyssey in heroic hexameter. The author says the examples look promising but notes that they cannot evaluate the Chinese translation directly.
Other links concern Jeremy Bentham’s reported cookbook for his Panopticon prison, the history of modern polyamory, and arguments about American suburbs and marriage. The author challenges a cited claim that only 18% of marriages are non-miserable, saying the source survey appears to show 61% of respondents describing their marriage as “very happy.” The post says other errors may also be present, though the author has not confirmed every discrepancy.
AI Moves Into Forecasting
The Metaculus result is the roundup’s clearest technology development: it offers an example of an AI system competing in a public forecasting contest and finishing ahead of human participants. If the reported ranking and development details are accurate, the result adds a concrete milestone to discussion about whether AI can make useful predictions across a range of questions.
It does not, by itself, show that AI systems are generally as good as expert forecasters. A competition result depends on the contest’s rules, the questions asked, how forecasts are scored and the number of participants. The post also describes one developer’s estimate of the time and cost involved; it is not an independently audited account of the system’s development. Readers should treat the win as a specific reported outcome, not a general measure of forecasting ability.
The roundup’s creative examples raise a separate issue: AI tools may help with forms of work, such as fitting a translation to poetic meter, that are difficult to do quickly by hand. But the post offers examples and personal impressions, not a controlled comparison of AI work with human translation or art. Its discussion is best read as a set of leads for further examination.
Popular right now
TCL QM64L 85-Inch QD-Mini LED 4K Smart Fire TV
Vitamix Propel Series 510 Professional-Grade Blender, 3 Programs, Black
Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Black
As an affiliate, we earn on qualifying purchases.
A Roundup, Not One Study
Astral Codex Ten’s “Links” posts collect outside writing, research and online examples, with brief commentary from the author. The October edition ranges across technology, culture and social questions rather than presenting a unified report. Many claims in it are summaries of material hosted elsewhere, so the original linked sources are needed to assess their full evidence and qualifications.
The author gives an explicit reliability warning: they say they have not independently checked every link and that commenters typically find evidence that around two or three links in each roundup are wrong or misleading. They say they correct items when they see problems and may highlight important changes later, but cannot guarantee all errors have been caught by publication. That caveat applies to the roundup as a whole, not only to the marriage discussion.
Several entries are openly framed as opinion or speculation. The author’s positive reaction to AI-assisted art and translation is a personal assessment. Likewise, the roundup presents competing views about marriage and suburbs rather than reporting a new policy or demographic finding. The clearest time-specific claim is the September Metaculus Cup result described in the October post.
“I haven’t independently verified each link.”
— Astral Codex Ten author, in the roundup’s opening note
Claims Still Need Checking
The roundup does not provide a complete independent audit of its links, and the author says some corrections may not yet have been made. The underlying Metaculus Cup article and The Economist’s account are the sources cited for Laertes’ win and development; the roundup alone does not establish the contest’s full scoring details, the bot’s methods or how its performance compares across other forecasting settings.
The marriage dispute also remains partly unresolved in the post. The author says they checked the “not too happy” figure but have not confirmed other alleged mistakes in the linked commentary. The roundup does not reproduce the survey data or a full analysis of the question wording. Its claims about AI art, translation and model behavior are also presented as examples, claims or impressions rather than verified conclusions about AI capabilities.
Check the Linked Sources
Readers seeking to verify the forecasting result can consult the Metaculus report and the Economist article cited in the roundup, including any details about the competition’s scoring and the winning system. FutureSearch is described as publicly available; the post says Mantic and Preseen can be accessed by contacting them for a waitlist or demonstration. It also links to a report that Mantic received $25 million in seed funding, though the roundup does not provide further funding terms.
For the other entries, the next step is to read the original research, commentary or examples rather than rely on the summaries alone. Astral Codex Ten says it may correct links and highlight important changes as errors are identified. No timetable for such updates is given, and the roundup does not announce a later report or follow-up event.
Key Questions
What is the October 2026 links post about?
It is an Astral Codex Ten roundup of links and commentary covering AI forecasting, AI-assisted art and translation, marriage, suburbs, polyamory and other subjects.
Did an AI bot win the Metaculus Cup?
The roundup reports that Laertes placed first in the September Metaculus Cup, citing The Economist and a Metaculus article. The post does not provide the full contest results or scoring details.
How reliable does the author say the links are?
The author says they have not independently verified every link and that commenters typically identify evidence that around two or three links per roundup are wrong or misleading. The author says corrections may follow.
Does the roundup prove AI can match human forecasters?
No. It reports one bot’s first-place finish in a specific competition. That result does not establish that AI systems generally match human forecasters across different questions or settings.
Source: astralcodexten.com
