Toby Ord argues in “Broad Timelines” for planning across uncertain AI timelines

Toby Ord argues in “Broad Timelines” for planning across uncertain AI timelines

Toby Ord’s “Broad Timelines” was published on March 9, 2026. Ord argues that lasting expert disagreement about when AI will have transformative effects calls for broad distributions of possible timelines, not choosing a single short or long estimate. He says that a named year can conceal how much uncertainty a forecast contains, distort public understanding, and mislead personal and policy planning. His proposed response is to act under uncertainty: hedge against transformative AI arriving early while also preserving work that could pay off if it arrives later. Ord argues that a longer timeline would mean a different political, technological and social setting, so plans tailored to present conditions may not fit. He also argues that projects taking years to build—such as founding organisations, developing research fields or writing books—can have substantial impact if there is time for their effects to compound. He concludes that the wider world should do more to prepare for early transformative AI, while people working to make the AI transition go well may be underweighting longer-term work. In either case, Ord says, “time is of the essence.” The article’s examples and forecasts express Ord’s reasoning about uncertainty, not known outcomes. The text includes references to charts, but their plotted details are not present in the retrieved text.

Publication: c18731bb-e44d-076a-a2be-4e2838319da4

attribution — pass: The page is a single essay by Toby Ord with his own argument, not a listing. The reading attributes the thesis, recommendations and conclusions to Ord and does not present them as established fact or as Escape Meridian's view.

March 9, 2026Toby Ord

summary — pass: The reading leads with Ord's central claim that experts disagree, so planners should hold broad rather than short or long timelines. It covers why single-year forecasts mislead, the hedging versus long-term-work balance, the different-world argument and the closing 'time is of the essence' line. Transformative AI is defined on the page. 'Broad distributions of possible timelines' is understandable in context. The question follows from Ord's closing call to funders, career advisors and movement builders.

You should have neither short timelines nor long timelines — but broad timelines.

date — pass: originalPublished is 2026-03-09, which matches the dateline 'March 9, 2026'. The page was modified on 2026-04-13 and ends with '19 March 2026', so a later revision is possible. However, the reading relies only on the essay's core argument, which is plainly part of the piece. It uses no later events and presents forecasts as Ord's reasoning, not as outcomes. This is a publisher page, so no archive disclosure is needed.

March 9, 2026Toby Ord

support — pass: Each claim matches a verbatim excerpt. The claim that the wider world should do more on early hedging, while the AI-transition community underweights long-term work, is supported by the conclusion. The communication and planning harms of single-year forecasts are stated in the source. The reading adds no unsupported benefits or costs.

The world at large is doing too little of the former, but those of us who care most about making the AI transition go well might be doing too little of the latter.

media — pass: No media was selected and no verified candidates were supplied. The reading correctly notes that the charts referenced in the text are not available for assessment.