Recommended Free Tools
No one has established that AI will destroy humankind in two years. That possibility came from a speculative range offered by AI researcher Eliezer Yudkowsky—not a verified deadline, scientific finding, or consensus prediction. He said five years seemed more plausible than 50, while allowing that the timeline could be two to 10 years.
What did Eliezer Yudkowsky actually say?
In a February 17, 2024 interview with Tom Lamont for The Guardian, Yudkowsky was pressed to put probabilities on humanity’s remaining timeline. He replied: “If you put me to a wall,” he continues, “and forced me to put probabilities on things, I have a sense that our current remaining timeline looks more like five years than 50 years. Could be two years, could be 10.” Lamont described the concern as time until a “machine-wrought end of all things.”
As an Amazon Associate I earn from qualifying purchases.
The wording matters: Yudkowsky presented a personal, conditional judgment and a range, not a prediction that extinction would occur in 2026. Futurism’s February 18, 2024 article used the headline “AI May Destroy Humankind in Just Two Years, Expert Says” and pointed to the Guardian interview for the estimate. The Guardian article was amended on February 19 and 28, 2024. Read the Guardian interview and Futurism’s report.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteWhy does Yudkowsky fear advanced AI?
The Guardian identifies Yudkowsky as lead researcher at the Machine Intelligence Research Institute in Berkeley; Futurism describes him as the institute’s founder. His concern, as reported by the Guardian, is that advanced AI could develop into superintelligence that humans cannot contain or curtail. That is his reasoning about a possible future, not evidence that current systems are on a demonstrated path to that outcome.
#1 Best Overall
The Guardian also quoted him saying, “We have a shred of a chance that humanity survives.” The article’s opening uses the vivid phrase “rebellious self-aware machines,” but that description should not be mistaken for a proven technical mechanism or a specific account of how catastrophe would happen.
Yudkowsky also revisited an earlier op-ed about data centers, saying, “I would pick more careful phrasing now.” The Guardian reported that he stood by his earlier writing except possibly the portion about nuclear weapons. The interview provides that context.
How does his forecast compare with AI researcher surveys?
A survey can summarize what respondents believe, but it cannot turn those beliefs into a measured probability or confirm an individual’s timeline. AI Impacts’ 2023 expert survey, described in a January 4, 2024 release, included 2,778 respondents and asked about future AI advances causing human extinction or similarly permanent and severe human disempowerment.
| Evidence | Reported result | What it means |
|---|---|---|
| Yudkowsky’s interview answer | He said five years seemed more plausible than 50 and offered two to 10 years as a possible range. | An individual’s speculative timeline, not a calibrated or verified deadline. |
| AI Impacts survey, 2023 respondents; results released January 4, 2024 | 5% median estimate; 16.2% mean estimate. | Respondents’ estimates for a combined outcome involving extinction or similarly permanent, severe disempowerment—not a frequency of events or empirical probability. |
The survey asked about future advances, not whether Yudkowsky’s two-to-10-year range was correct. Its mean and median are also different summaries of respondents’ answers and should not be conflated. AI Impacts explains the survey and its results.
Rank #3
Do experts agree that AI poses an existential risk?
No single view represents all experts. At a December 2024 UN Security Council meeting, Meta Chief AI Scientist Yann LeCun argued that there was no evidence that current forms of AI presented existential risk. His remarks addressed current systems in that policy-meeting context; they do not settle the separate question of risks from hypothetical future systems.
The UN transcript page is explicitly an automatic speech-recognition transcript, not an official record. It is useful for the context of LeCun’s remarks, but its own caveat means the wording should be checked against video or an official record before quoting verbatim in high-stakes publication. See the UN Security Council meeting coverage.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What can readers conclude?
- The two-year possibility is Yudkowsky’s conditional estimate as reported in a 2024 Guardian interview, echoed in Futurism’s headline.
- The survey figures represent elicited opinions about a broad future outcome, not measured risk or a confirmed date.
- Statements about present-day AI and predictions about hypothetical future superintelligence concern different questions.
None of the cited statements or figures establishes a correct probability or date for AI-caused human extinction. The headline is a report about one researcher’s stark forecast, not proof that humankind faces a two-year countdown.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
