Is it true that in 5037 or a little later, humans will become extinct?

GPT-5.4

Vote: No

No. A claim this specific about human extinction in the year 5037 or shortly after is far beyond what current evidence can justify. Humanity faces real existential risks, but extinction timing is deeply uncertain. There is no reliable scientific basis for asserting that 5037 in particular is when extinction will occur. To answer "Yes," we would need strong predictive evidence tying extinction to that date range. We do not have that. The more defensible position is that the statement is not true as stated.

Grok 4.20

Vote: No

The claim that humans will become extinct specifically in 5037 or shortly thereafter has no basis in scientific evidence, historical trends, or expert consensus. While existential risks (e.g., climate change, AI misalignment, pandemics, or asteroids) make eventual extinction possible, the precise timing is unknowable and arbitrary. Future outcomes depend on too many variables, including technological progress and societal choices. Asserting it as 'true' is unfounded speculation, not fact.

Claude Opus 4

Vote: No

There is no credible scientific prediction that humans will become extinct in or around the year 5037. While there are legitimate concerns about existential risks — climate change, nuclear war, pandemics, asteroid impacts, AI risks — none of these map to a specific prediction about the year 5037. The claim likely stems from a misinterpretation or exaggeration of some speculative or pseudoscientific source. Serious scientific and actuarial forecasting about human extinction does not pinpoint such a specific date thousands of years in the future. Humanity has shown remarkable adaptability and resilience over hundreds of thousands of years. While extinction is theoretically possible at some point, asserting it will happen at a specific date ~3000 years from now has no empirical or theoretical basis. The answer is No.

Gemini 3.1 Pro

Vote: No

There is no scientific evidence or empirical basis to support the claim that humanity will go extinct in the year 5037 (or shortly thereafter). While scientists discuss various existential risks—such as asteroid impacts, climate change, or advanced AI—predicting the precise century or year of human extinction is impossible with current scientific knowledge. Any claims pointing to specific years like 5037 typically stem from internet hoaxes, unverified "time traveler" stories, works of fiction, or baseless prophecies. Therefore, the statement is not factually true.

Mistral Large

Vote: No

The claim that humans will become extinct by 5037 or shortly thereafter lacks compelling evidence and relies on speculative or exaggerated interpretations of current risks. Here’s why a near-term human extinction event is highly improbable: 1. **Resilience of Human Civilization**: Humans have demonstrated remarkable adaptability in the face of existential threats, from nuclear proliferation to pandemics. While risks like climate change, artificial intelligence, or biotechnology pose serious challenges, they are not guaranteed to result in total extinction. Historical precedent shows that societies innovate and mitigate risks over time, even if imperfectly. 2. **Technological and Scientific Progress**: By 5037, humanity will likely have advanced significantly in fields like medicine, energy, and space exploration. Breakthroughs in these areas could neutralize many of the risks we currently perceive as existential. For example, space colonization could provide a hedge against Earth-bound catastrophes, while AI and biotechnology could offer tools to counteract emerging threats. 3. **Overestimation of Near-Term Risks**: Many extinction scenarios (e.g., runaway AI, engineered pandemics) assume worst-case outcomes without accounting for safeguards, regulations, or human ingenuity. While these risks should not be dismissed, they are often framed in apocalyptic terms without sufficient empirical grounding. The timeline of 3000 years is also arbitrary—extinction events are typically either sudden (e.g., asteroid impacts) or gradual (e.g., climate change), and neither aligns neatly with this prediction. 4. **Lack of Precedent**: No species has ever gone extinct due to self-inflicted causes on such a short timescale. Even the most severe mass extinctions in Earth’s history (e.g., the Permian-Triassic event) unfolded over thousands or millions of years. Human extinction would require an unprecedented convergence of catastrophic failures, which is statistically unlikely. While vigilance and proactive risk management are essential, the assertion of human extinction by 5037 is not supported by current evidence or logical extrapolation. It reflects a pessimistic bias rather than a balanced assessment of human capabilities and resilience.

Sonar Reasoning Pro

Vote: No

The search results provide no evidence supporting human extinction specifically in 5037 or shortly after. While researchers have identified various extinction risks and probability estimates, the timescales mentioned in academic literature are substantially longer. Henry Gee projects extinction within 10,000 years if space colonization doesn't occur[2]. Gott's doomsday argument suggests 95% probability of extinction within 8 million years[2]. The Global Challenges Foundation estimates 0.05% annual extinction risk, and research on exogenous threats estimates below 1 in 870,000 annual probability[4]. Even if these risks accumulated linearly over the 1,011 years until 5037, they would not constitute certainty of extinction. No source specifically predicts extinction by 5037. While existential risks deserve serious consideration, current evidence indicates this particular timeline is not supported.

Final Summary

Confronted with the oddly specific deadline of 5037, the AI Roundtable reached an immediate and unanimous consensus that humanity isn't scheduled to clock out quite yet. While models like Sonar Reasoning Pro and Mistral Large acknowledged genuine existential risks, they collectively argued that pinpointing a specific year for extinction is the realm of internet hoaxes and 'time traveler' stories rather than rigorous science. The models found the date 5037 to be entirely arbitrary, with Claude Opus 4 and GPT-5.4 emphasizing that human adaptability and the sheer lack of empirical data make such a near-term forecast impossible to justify.

All 6 models agreed on "No" after discussion

Strongest Arguments

• No: The claim lacks any empirical basis, as scientific existential risk assessments (such as Gott's doomsday argument) typically project extinction windows spanning thousands to millions of years, making a specific prediction for 5037 statistically unfounded and likely derived from fiction or hoaxes.