Over the past half-century, Americans have been told that climate change wasn't merely an environmental problem. It was a coming apocalypse.
We were given ultimatums. We were shown maps of major cities that would soon be underwater. Demagogues pounded the table and declared an “existential threat.” Children were traumatized into believing their planet was being destroyed.
Governments reorganized energy policy around the assumption that catastrophe was approaching unless humanity rapidly abandoned fossil fuels.
Now, we have a big change in the climate of these pseudo-scientific opinions.
The next generation of scenarios intended to inform climate research for the UN Intergovernmental Panel on Climate Change has dramatically lowered the ceiling.
The new CMIP7 scenarios released September 1 project a range of approximately 1.6 degrees Celsius to 3.3 degrees Celsius of warming by 2100. The highest scenarios used in the previous three generations of climate modeling produced roughly 4.6 to 4.9 degrees.
The scientific “consensus” has just collapsed.
Carbon Brief, hardly a bastion of climate skepticism, describes it plainly: This is the first time in four generations of climate modeling that the top of the scenario range has fallen.
The new high scenario contains roughly half the cumulative carbon dioxide emissions of SSP5-8.5. Even more strikingly, the entire “high” tier from the previous generation now sits above anything in the new CMIP7 scenario set through 2100.
And that creates a rather awkward question. What happened to the apocalypse?
Michael Shellenberger has taken that question directly to the climate establishment in a new investigation at Public.
For decades, climate scientists said climate change was apocalyptic. It never was, as new UN scenarios prove. Scientists knew this and pushed a falsified scenario anyway. Now exposed, the scientists are falsely claiming that it was their policies that averted catastrophe. pic.twitter.com/og9t7CLPoz
— Michael Shellenberger (@shellenberger) September 3, 2026
His charge is considerably more serious than simply saying the models changed. He argues that scientists continued treating an implausibly extreme emissions scenario as a meaningful future long after evidence had accumulated that its underlying assumptions weren't realistic — and that some are now claiming climate policies deserve credit for preventing the future that critics said wasn't plausible in the first place.
That's quite a trick. For years, skeptics of climate alarmism were told that dismissing RCP8.5 as unrealistic amounted to minimizing climate change. Now prominent climate scientists themselves are acknowledging that the scenario is no longer plausible.
Chris Smith, one of the researchers involved with the new scenarios, recently put it this way: “Back in 2010, RCP8.5 wasn’t an implausible future, we’ve now made it an implausible future.”
NASA Goddard Institute for Space Studies director Gavin Schmidt similarly argued that scenarios changed “in light of what we (as a society) are doing” and that “‘business’ is no longer ‘usual.’”
IPCC Working Group I co-chair Robert Vautard pointed to climate policy, arguing that the old high scenarios assumed no such policies while today “there ARE now many climate policies in many countries,” particularly following the Paris Agreement.
Shellenberger's argument is that this amounts to rewriting the history of the debate.
The problem with RCP8.5 wasn't simply that humanity eventually enacted enough climate policy to escape it. Critics had been pointing out for years that the scenario required some extraordinary assumptions about energy consumption — most notoriously an enormous increase in coal use.
Researchers Justin Ritchie and Hadi Dowlatabadi zeroed in on this problem back in 2017. RCP8.5 effectively treated coal as an enormously abundant backstop fuel, producing a future in which global coal consumption exploded rather than following the technological and economic transitions already underway.
The result was an emissions pathway that could generate spectacular warming numbers. It was subsequently used in thousands of studies, government assessments and countless media stories as a representation of what could happen without sufficiently aggressive climate action.
There is a major distinction worth making here. The IPCC itself does not design the CMIP scenarios; it assesses scientific literature and scenarios produced by the broader research community. The World Climate Research Programme's CMIP process coordinates the modeling work that feeds heavily into IPCC assessments.
But that institutional distinction doesn't make the political problem disappear.
The extreme scenarios became deeply embedded in the public discussion of climate change. Worst-case projections migrated from exploratory scientific scenarios into headlines, government risk assessments and political campaigns. A scenario designed to explore an extreme emissions pathway could easily become, by the time it reached the public, a picture of where humanity was headed unless governments acted immediately.
And the historical record of environmental apocalypse predictions doesn't exactly inspire confidence.
Over the past half-century, the public has been warned about impending global cooling, mass famine, resource exhaustion, disappearing Arctic ice, submerged cities, tens of millions of climate refugees and various deadlines to “save the planet.”
The details change, but the political rhythm remains remarkably familiar: catastrophe is approaching, experts have calculated the deadline, and enormous economic or political changes must happen. When the actual data come in refuting the climate model, the timeframe is simply moved.
The apocalypse doesn’t get called off, it just gets moved to a future data long into the future; preferably, a date when nobody could ever be held accountable.
Now, the new scenarios make the distinction between climate science and climate alarmism more important than ever.
And there is another fascinating development occurring at exactly the same time.
After decades of politicians insisting that energy consumption had to be restrained and fossil-fuel generation rapidly displaced, the technological economy suddenly needs staggering quantities of electricity.
Artificial intelligence happened. The data-center boom is now forcing governments, utilities and technology companies to confront energy abundance as an economic necessity rather than an environmental vice.
The Electric Power Research Institute estimates that data centers could consume 9 to 17 percent of all U.S. electricity by 2030, up from roughly 4 to 5 percent today. Its updated 2026 projection for data-center electricity demand is 60 percent higher than its previous scenarios. Under its reference-policy assumptions, natural gas supplies the largest share of incremental generation.
Globally, the International Energy Agency projects data-center electricity consumption will roughly double by 2030, with AI-oriented accelerated servers accounting for nearly half the increase.
Texas provides a glimpse of the scale involved. Electricity demand is surging after years of relative stability, with data centers expected to account for a huge portion of new peak demand. The flood of proposed projects has become so enormous that Texas is now auditing data-center interconnection requests to separate real projects from speculative “ghost demand.”
That doesn't prove the new climate scenarios were rewritten because of AI or data centers. The CMIP7 process predates the latest stage of the AI boom, and the scenario designers point to a broader combination of technological change, energy economics, existing climate policy and updated emissions trends.
But politically, the timing is impossible to miss.
The West spent years constructing an energy policy around scarcity just as it entered a technological competition that demands abundance.
Suddenly, shutting down reliable power plants doesn't look quite as fashionable when hyperscale data centers need gigawatts.
Natural gas is becoming strategically important again. Nuclear power is enjoying a renaissance. Technology companies that spent years advertising their green credentials are scrambling to secure enormous blocks of reliable electricity.
The material needs of the AI revolution are colliding with the ideology of climate austerity.
Something had to give. And it looks like it just did.
