A sourced scoreboard of the AI industry’s dated predictions.

"85% Of People Will be Unemployable. According to my models, only 15% of working age adults will be employed once AI and robotics take off."

"Both methods point to roughly 200,000 to 300,000 U.S. jobs displaced or foregone due to AI in 2025."

"In 2026 at least one — and probably several — companies (Figure, Tesla Optimus, Unitree, Boston Dynamics Atlas, others) will cross into genuine product-market fit (PMF)." Shapiro called 2026 the "year of the robot."

"I'm updating my predictions. At the end of 2024 I said: 2025 — the year of the agent (so far, quite true); 2026 — the year of the robot; 2027 — the year of superintelligence. I now think that superintelligence will also happen in 2026."

"2025 has seen the arrival of agents that can do real cognitive work… 2026 will likely see the arrival of systems that can figure out novel insights… 2027 may see the arrival of robots that can do tasks in the real world."

AI could wipe out half of all entry-level white-collar jobs and spike unemployment to 10–20% in the next one to five years.

"I think we'll be there in three to six months — where AI is writing 90 percent of the code. And then in twelve months, we may be in a world where AI is writing essentially all of the code."

"Possibly by 2026 or 2027 (and almost certainly no later than 2030)… a ‘country of geniuses in a datacenter.’"

"We believe that, in 2025, we may see the first AI agents ‘join the workforce’ and materially change the output of companies."

For 2025: no AGI; under ~10% of the workforce replaced by AI; driverless cars still geographically limited; possibly no community-consensus ‘GPT-5-level’ leap.

"At the end of 2024 I said that 2025 would be the year of the agent and 2026 would be the year of the robot. I hold to that. 2027 is the year of superintelligence."

"Why I'm quitting AI… Mostly I'm just quitting AI on YouTube and I'm quitting the content mill… I'm going to wrap up all my AI-related projects."

"AI-enabled biology and medicine will allow us to compress the progress that human biologists would have achieved over the next 50–100 years into 5–10 years" — the “compressed 21st century.”

"It is possible that we will have superintelligence in a few thousand days (!); it may take longer, but I'm confident we'll get there."

"Maybe I was right, and we achieved AGI in September of 2024? We need to wait for the benchmarks, but we just had a major step in that direction."

"I have predicted superintelligence by 2027 and a technical academic AGI in about 2025… Real deal AGI I expect to be later in 2025 through the end of 2026."

"$1 million (later raised to $10M) says" AI will not be smarter than any individual human by the end of 2025.

By end of 2024: many GPT-4-level models, no massive advance (no GPT-5, or a disappointing one), price wars, little moat, no robust fix for hallucinations, and only modest corporate adoption and profits.

"Earlier this year I predicted that we would have AGI within 18 months — that was March of 2023, so that means my prediction was by September 24, 2024 we would have AGI. I am here to reaffirm that prediction. We will have AGI within 12 months."

"In 18 months any possible definition of AGI that you have will be satisfied." Shapiro framed this as AGI arriving within 18 months of March 2023 — i.e. by roughly September 2024.

Seven predictions about GPT-4: it would still hallucinate, remain unreliable at physical/psychological/mathematical reasoning, not give reliable medical advice, and not be safely hooked up to downstream programs.

"Deep learning is hitting a wall" — scaling makes models more fluent but not more trustworthy, with diminishing returns on reasoning, truthfulness and common sense.

"Imagine a world where, for decades, everything — housing, education, food, clothing, etc. — became half as expensive every two years."