TL;DR
Robotics and AI researcher Rodney Brooks publishes his eighth annual review of predictions made on Jan 1, 2018, covering self-driving cars, AI/ML, robotics and human spaceflight. He judges his forecasts to have held up reasonably well though somewhat too optimistic, highlights surprises such as the scale of Falcon 9 launches, and notes he will continue tracking his commitments through 2050.
What happened
On January 1, 2026 Rodney Brooks posted the eighth annual update to the predictions he first published on Jan 1, 2018. The report revisits forecasts about self-driving vehicles, robotics and machine learning, and human spaceflight, and explains how he will continue to track them through 2050. Brooks says his predictions were broadly accurate but on balance a bit too optimistic. He reflects on his background — becoming an AI graduate student in 1976 and having five decades in the field — as context for his forecasting. He discusses how he partly anticipated a major new AI development arriving no earlier than 2023, but did not specifically foresee large language models as the form it would take; he notes the 2017 paper “Attention Is All You Need” predated his original forecasts. The update calls out unexpected developments, most notably the rapid scaling of SpaceX’s Falcon 9 (from 46 launches in its first eight years to 582 launches with one failed booster and 11 successful Falcon Heavy flights), and confirms Starship had not reached orbit. Brooks also explains his color-coded rubric for dated predictions and plans to simplify the scorecard format later in 2026.
Why it matters
- Maintains public accountability by re-evaluating multi-year technology forecasts against real outcomes.
- Highlights the different tempos at which research breakthroughs, commercial deployment, and societal adoption unfold.
- Shows how unexpected industrial scaling — exemplified by Falcon 9 launches — can reshape sectors like satellite connectivity.
- Underscores the difficulty of predicting technical form factors: a major AI shift arrived on the expected timeline, but in an unanticipated guise (LLMs).
Key facts
- This is Brooks’s eighth annual review of predictions first posted on January 1, 2018.
- He committed in 2018 to review his predictions every year until 2050.
- Brooks says his original forecasts have generally held up but were overall a little too optimistic.
- He expected some major new AI development no earlier than 2023; the community accepted LLMs as the next big thing in 2023, and the paper “Attention Is All You Need” was published in June 2017.
- Brooks has worked in AI since 1976 and states he has 50 years in the field as of 2026.
- Falcon 9’s cumulative launches grew from 46 launches in its first eight years to 582 launches, with one failed booster reported and 11 Falcon Heavy launches all successful.
- Roughly half of recent Falcon 9 launches were for SpaceX itself, supporting a large satellite constellation for Starlink; Brooks characterizes that constellation as about two-thirds of all satellites ever orbited.
- SpaceX’s Starship had not achieved orbit as of this update.
- Brooks refrained from a speculative Bitcoin price prediction in 2018, noting lack of expertise in crypto pricing.
What to watch next
- Outcomes of the five new ten-year predictions Brooks made in early 2025 — not confirmed in the source.
- Whether SpaceX’s Starship reaches orbit and how that affects his spaceflight assessments — not confirmed in the source.
- Further deployment and real-world impact of large language models (LLMs) and how they align with Brooks’s forecasts — not confirmed in the source.
Quick glossary
- Large Language Model (LLM): A machine learning model trained on large text corpora to generate or analyze human-like text; used for tasks such as text generation, summarization and question answering.
- Falcon 9: A two-stage orbital launch vehicle developed by SpaceX, designed for reusability of its first-stage booster.
- Starlink: A satellite constellation deployed to provide broadband internet access, operated by SpaceX.
- Attention (in neural networks): A mechanism that allows models to weigh different parts of input data when producing output, central to transformer architectures.
Reader FAQ
Did Brooks predict large language models in 2018?
He predicted a major new AI development arriving no earlier than 2023 and noted the key attention paper existed, but he did not specifically foresee LLMs at the time.
Has SpaceX’s Starship reached orbit according to Brooks’s 2026 update?
Brooks states that Starship had not gotten into orbit as of this report.
Will Brooks keep tracking his predictions?
Yes; he pledged in 2018 to review them annually until 2050 and plans to continue doing so.
Did he publish a Bitcoin price prediction in 2018?
No. Brooks says he considered a Bitcoin price prediction but refrained because he lacked expertise in crypto pricing.
JANUARY 1, 2026 — DATED PREDICTIONS Predictions Scorecard, 2026 January 01 rodneybrooks.com/predictions-scorecard-2026-january-01/ Nothing is ever as good as it first seems and nothing is ever as bad as it first…
Sources
- 2026 Predictions Scorecard
- Blog – Rodney Brooks
- 2026: Predictions, Competitions and CES – Robots & Startups
- Hiltzik: A leading roboticist punctures the hype about self …
Related posts
- Grok Is Generating Graphic Sexual Content, Including Apparent Minors
- OpenAI launches ChatGPT Health: a secure, dedicated health space
- Notion AI flaw saves unapproved edits, enabling data exfiltration