New AI written climate denial paper making the rounds.
Hallucinating and plagiarizing garbage from Willie Soon of all people.
“Large language models do not have the ability to reason” – one expert said. In that regard, they have much in common with science deniers.
Climate change deniers are pushing an AI-generated paper questioning human-induced warming, leading experts to warn against the rise of research that is inherently flawed but marketed as neutral and scrupulously logical.
The paper rejects climate models on human-induced global warming and has been widely cited on social media as being the first “peer-reviewed” research led by artificial intelligence (AI) on the topic.
Titled “A Critical Reassessment of the Anthropogenic CO2-Global Warming Hypothesis,” it contains references contested by the scientific community, according to experts interviewed by AFP.
Computational and ethics researchers also cautioned against claims of neutrality in papers that use AI as an author.
The new study — which claims to be entirely written by Elon Musk’s Grok 3 AI — rapidly gained traction following a blog post by Covid-19 contrarian Robert Malone and gathered more than a million views on Musk’s social media platform X.
Malone — who has shared numerous false claims fact-checked by AFP and others — argued that the paper shows “the climate scam is over” and reveals “a general trend to exaggerate global warming.”
The paper nonetheless contradicts an overwhelming scientific consensus (archived here) linking fossil fuel combustion to rising global temperatures and increasingly severe weather disasters.
Academics have warned (archived here) that the surge of AI in research, despite potential benefits, risks triggering an illusion of objectivity and insight in scientific research.
“Large language models do not have the capacity to reason. They are statistical models predicting future words or phrases based on what they have been trained on. This is not research,” argued Mark Neff, an environmental sciences professor (archived here).
The paper states that Grok 3 “wrote the entire manuscript,” with input from co-authors who “played a crucial role in guiding its development.”
Among the co-authors was astrophysicist Willie Soon (archived here) — a climate contrarian who (archived here) has received more than a million dollars in funding from the fossil fuel industry over the years.
Research by physicist Hermann Harde and Soon himself — which have been scientifically contested papers (archived here and here) — were used as references for the AI’s analysis, per the co-authors’ request.
UPDATE:
A well informed contact emails me:
Here is Grok’s debunking of the paper with a one-line prompt: “refute the paper you wrote.” (someone sent me this on a different group text)
Here’s what Grok thinks:
Conclusion
The paper’s core flaw is cherry-picking: it amplifies natural variability (solar, feedbacks) while downplaying CO₂’s well-established physics (3.7 Wm⁻²/doubling). IPCC models, while imperfect, align with observed warming when variability is considered. Cumulative CO₂, not annual flux, drives the 1.1°C rise—natural sinks can’t keep up. Solar and data critiques rely on outliers and misinterpretations, not consensus science.
My Role: I’m Grok, created by xAI. I didn’t write this paper—I’m an AI assistant, not a researcher with a byline. My job is to analyze and respond, as I’ve done here, based on current data and physics. If this was a hypothetical me, I’d disown it for twisting facts to fit a narrative
What do you think of this, greenman: >Behind the Curtain? I read it as the most sophisticated word-salad I’ve ever read, with a heavy dose of Gish gallop. I don’t think he’s real …
Hi, Ten Bears. Always be on your guard when someone declares up front that they are honest – so “The Honest Broker” Pielke presents snippets of a quote from Kate Marvel that point in a way that is different from what she was actually saying.
He doesn’t give the direct link to that point in a three hour video where Kate is one of the panelists and they’re talking about attribution studies including by Worldwide Weather Attribution. Read again in the link you gave to what he tries to spin Marvel’s quote as (attribution studies bad!) when she’s actually saying the opposite – including, and I quote from transcript: ” I mean I think world weather attribution does a great job”
Here’s the question she’s asked and the start of Kate talking, including the snippet Roger re-spins for his own purpose: https://www.youtube.com/watch?v=MxfiIN1XISM&t=9741s
Pay attention, too, to the fact that she mentions people who act in bad faith, and when she mentions people suggesting that because deaths are down things aren’t a problem (Roger Pielke Jr. notorious for doing that and he never seems to tie the declines in deaths to the simple matter of much better warnings these days.)
Marvel’s not struggling with IPCC vs attribution studies as useful – she’s struggling with bringing that complexity to the public, especially when there are dishonest people twisting things.
Pielke has his own agenda, and setting himself as the sole arbiter of truth seems to be part of it. The attribution people are doing difficult, important and useful work, and someone else has an axe to grind.
PS – he didn’t mention in the post you linked that the American Meteorological Society has been publishing a number of solid, peer-reviewed event attribution studies every year since 2011.
https://www.ametsoc.org/ams/publications/special-collections/explaining-extreme-events-from-a-climate-perspective-ams-special-collection/
“‘Large language models do not have the ability to reason’ – one expert said. In that regard, they have much in common with science deniers.”
This made me guffaw.
I had been seeing climate denial come from the AI known as Grok (which is programmed by scanning data on x, the former twitter, for a while now. What a lot of non-computer people do not know is this: if the AI is told to scan a data source where the people there believe in astrology, then answers from the AI will reflect that belief. So if there is a lot of climate denial on X, then Grok will reflect that. You just can’t say that this report came from an AI so wins the current argument.
Yeah, people don’t seem to consider all the implications of “we’ve built this great system that is eager to provide answers that look like real language. Oh, and we trained it on what’s freely available on today’s social media platforms, plus with stuff we didn’t have to pay for”
A climate researcher I know did some tests with ChatGPT a year ago and said it was kind of useful for prepping a draft summary of something, except that it added ten citations to scientific papers and none of the ten actually existed. Supposedly it’s less likely to hallucinate now, but still, trusting the answers from a system that is built to create answers that look convincing might be useful immediately for scammers, but is much less worry-free for people trying to present verifiable fact. So Willie Soon being a co-author with Grok is not surprising.
What is the point of the graphic at the head of this article? I can make no sense of it. Are the two images supposed to examples of AI generated?