Who called it? AI predictions, checked.

“…next year for sure, we'll have over a million robotaxis on the road. The fleet wakes up with an over the air update; that's all it takes.”

No Tesla robotaxi service ran in 2020. Paid rides began in Austin in June 2025 with safety monitors on board, and by mid-2026 the driverless fleet was about 20 cars.

Due Dec 2020Adoption

“They should stop training radiologists now.”

Radiologists are still trained and hired. AI now helps read scans, yet Mayo Clinic's radiology staff alone grew 55% after 2016, to more than 400. Consults, hard cases, and sign-off stayed human.

Due Dec 2021Jobs

“I think we will be there in three to six months, where AI is writing 90% of the code. And then, in 12 months, we may be in a world where AI is writing essentially all of the code.”

Six months on, there was no sign of AI writing 90% of code across the software industry. Measuring who writes code is hard, and a narrow reading (code at AI labs) would score differently.

Due Sep 2025Adoption

“We believe that, in 2025, we may see the first AI agents 'join the workforce' and materially change the output of companies.”

Coding agents spread fast in 2025 and some firms rebuilt workflows around them. Evidence that agents materially changed company output at scale stayed thin, and "may" leaves room either way.

Due Dec 2025Adoption

In 2025, Meta and other companies would probably have an AI that can effectively work as a midlevel engineer who writes code (paraphrased from the Joe Rogan Experience #2255).

Coding agents took on more routine engineering work, but Meta's next step was to set targets for how much code its engineers produce with AI, not to replace midlevel engineers.

Due Dec 2025Capability

A computer would need at least another ten years to beat one of the world's elite Go professionals.

AlphaGo beat European champion Fan Hui in October 2015, then Lee Sedol, one of the world's best, 4 to 1 in March 2016: about a decade early.

Due Dec 2025Capability

The median expert expected AI to reach gold-medal level at the International Mathematical Olympiad around 2030; the median superforecaster, around 2035.

AI systems reached gold-medal level at the July 2025 Olympiad, five years before the median expert forecast and ten before the median superforecaster's. FRI finds forecasters have repeatedly underestimated progress on benchmarks.

Due Dec 2030Capability

“…a superhuman coder (SC): an AI system that can do any coding tasks that the best AGI company engineer does, while being much faster and cheaper.”

Not yet due.

Due Mar 2027Capability

“AGI by 2027 is strikingly plausible.”

Not yet due.

Due Dec 2027Capability

“I could easily see 30 percent of that getting replaced by AI and automation over a five-year period.”

Not yet due.

Due May 2028Jobs

“In the next five years, this will change completely. You won't have to use different apps for different tasks. You'll simply tell your device, in everyday language, what you want to do.”

Not yet due. Assistant and agent products multiplied after 2023, but not yet the everyday replacement for apps he described.

Due Nov 2028Adoption

“If we specified AGI to be something very specific, a set of tests where a software program can do very well — or maybe 8% better than most people — I believe we will get there within 5 years.”

Not yet due. With a test-based definition, much depends on which tests count.

Due Mar 2029Capability

A machine will pass a rigorous Turing test by 2029. He staked $20,000 on it against Mitch Kapor in a formal Long Bet, and restated the date in 2024.

Not yet due. The bet's long, formal interviews have not been held.

Due Dec 2029Capability

“…they could drive a 7% (or almost $7 trillion) increase in global GDP and lift productivity growth by 1.5 percentage points over a 10-year period.”

Not yet due. Note that exposure is not loss: the 300 million figure counts jobs with tasks AI could do, not jobs that will disappear.

Due Dec 2033Economy

“…these macroeconomic effects appear nontrivial but modest—no more than a 0.66% increase in total factor productivity (TFP) over 10 years.”

Not yet due. Productivity statistics can't yet separate AI's contribution from everything else.

Due Dec 2034Economy