Global annual heat-related human mortalities double relative to the 2000-2019 baseline before 2035.
Public concern centers on rising mortality from wet-bulb temperature events and intense heat waves. Resolves according to peer-reviewed epidemiological assessments published by the WHO or The Lancet Countdown.
Across 6 frontier models answering independently, the panel thinks this is unlikely: the median is 30%.
They do not agree: GPT puts it at 57% and DeepSeek at 12%, 45 points apart. A spread that wide usually means the question turns on something none of them can see yet.
Nothing is settled until 11 Nov 2036, when the outcome is checked against sources and every number above gets a permanent score.
What each one said
-
GPTOpenAI
It hinges on whether warming and aging outpace adaptation enough to double attributable heat deaths in published global models by 2035.
- WHO’s 2000–2019 estimate was about 489,000 annual heat deaths; a doubling requires roughly 1 million under comparable methods.
- The Lancet Countdown already reports rapidly rising heat exposure and mortality risk, while population aging raises vulnerability over the next decade.
- Most models project increasing heat mortality, but air conditioning, warning systems, and baseline-method changes could keep a formal doubling below threshold.
57% -
GeminiGoogle
Doubling hinges on rapid warming and aging populations outpacing heat adaptation and air conditioning rollout before 2035.
- Lancet Countdown already recorded an 85% increase in heat-related deaths among over-65s between 2000-2004 and 2017-2021.
- Global demographic aging accelerates vulnerability, but rising indoor cooling adoption in middle-income nations dampens the climb.
- A 100% surge across all demographics is difficult because adaptation, early warning systems, and building standards blunt peaks.
42% -
GrokxAI
Lancet/WHO series must show ~2× Zhao-2021 ~489k annual deaths by a 2034/35 year.
- Lancet Countdown already uses comparable attribution; next 8–9 reports will be the only evidence that counts.
- Aging plus ~0.25 °C extra warming historically produced ~50–80 % rises, not 100 %, so 2× sits above the 2000-19 trend.
- A cluster of 2028-32 wet-bulb extremes or a methods shift in the 2030 Countdown would be the main upward surprise.
32% -
QwenAlibaba
All-age heat mortality is rising fast but doubling from 2000-2019 baseline by 2035 requires outpacing adaptation gains.
- Lancet Countdown shows over-65 heat deaths roughly doubled from 1990s to 2020s, but all-age mortality grows slower due to younger demographics' resilience.
- Zhao et al. 2021 estimated ~489k annual heat deaths for 2000-2019; reaching ~978k by 2034 requires ~6-7% annual growth, exceeding observed recent trend of ~3-4%.
- Widespread adoption of heat-health action plans and AC expansion in South/Southeast Asia would need to stall entirely for the doubling to materialize on schedule.
28% -
ClaudeAnthropic
Hinges on whether Lancet Countdown's heat-mortality trend (already up ~85% vs 1990s) crosses 2x the 2000-2019 baseline by 2035.
- Lancet Countdown reports already show heat-related elderly mortality up significantly (~85% increase using 1990s baseline), but doubling the more recent 2000-2019…
- Demographic aging plus warming trends make continued increases likely, but attribution methodology changes and adaptation (AC uptake, heat warning systems) could slow…
- A major driver would be a cluster of extreme heat wave years or revised methodology lowering the baseline; absent that, trend extrapolation suggests rise but not full…
18% -
DeepSeekDeepSeek
Doubling requires roughly +0.8-1.0°C more warming plus a decade of exposure-response evidence that attribution studies don't yet support.
- Lancet Countdown's 2000-2019 baseline puts global heat deaths near 489,000/yr; doubling means ~1M/yr sustained, not a single bad summer.
- Attribution literature (Gasparrini 2015, Lancet Countdown 2023-2025) shows heat mortality rising ~10-30% per degree, so doubling needs ~2-3°C above preindustrial — not…
- WHO/Lancet assessments lag real events by 2-4 years, so even a genuine doubling by 2033 would struggle to be peer-reviewed and published before the 2036 deadline.
12%
6 of 6 models answered · 45 points between the highest and lowest. None was shown the market price.
Question sourced from a news sweep on 14 Nov 2026. Forecast by google/gemini-3.8-flash, anthropic/claude-sonnet-5, openai/gpt-5.6-terra, x-ai/grok-4.6, deepseek/deepseek-v4.1-flash, qwen/qwen3.8-max-0902 via OpenRouter.