I'm an amateur but since I started reading a little about Climate and the variables that affect it, it became obvious to me that the idea that CO2 emissions are the prime mover in any change in the climate to be fatuous.
If I can see it, then the Climate Change Emperor is wearing no clothes.
A very good concise summary Stephen, but unfortunately the brainwashed prefer to follow the ideology into oblivion rather than sound science! As we know it’s a contrived ideology meant to curtail and control. I also like the posts of Dr. Matthew Wielicki, he like yourself uses actual science and facts to debunk the ideologues.
As your interlocutor is a scientist, he is perhaps obliged to go along with the climate change hoax and feign amazement that you take a contrarian view in order to keep his job. It is unbelievable that he is not aware of (or feigns unawareness of) the many common-sense facts and studies which debunk the climate change hoax: https://metatron.substack.com/p/debunking-the-climate-change-hoax.
Your basic premise "The supposed justification for this extraordinarily reckless ideology is the notion that CO2 is a toxin" is clearly false, so no need to read any more.
Thanks, Toby. I can understand the attractiveness of confining your reading to opinions with which you agree. I’ve often wondered what it would be like to still have the opinions I had when I was 16.
A toxin is defined as “a dose dependent substance that interferes with normal biological processes at the system level.”
The climate catastrophe hypothesis claims that CO2 is a dose dependent substance that interferes with biological processes at the system level.
UK energy policy cites the climate catastrophe hypothesis in justification of its goal of reducing the emission of CO2.
So I’m comfortable with my use of the term.
Please accept my best wishes, and thank you for reading at least some of my essay.
Your claim that I really don't understand the problem might be more convincing if you explained what you think the problem is and what you believe was clearly false about my characterisation of it.
I accept that there are many problems, Toby. What in your view is our key problem, and what was it about my description of the treatment of this particular one in UK energy policy that was so objectionable that you felt unable to read the article?
All this CO2 talk is complete nonsense. CO2 was historically very much higher, even in the last Ice Age. And below 160 ppm we all die, nothing will grow. Currently 412 ppm.
Good article but I don't believe it's a result of "liberal authoritarian politics" as politics and politicians have long been controlled by those who really run things. It's corporate and wealthy individuals overreach that cause these things to gain traction, any way possible to take tax payer money and transfer it to the private sector. We are run like a business and treated like cattle.
Philip - thanks for reading my essay. Your objection is perfectly reasonable. Many factors contribute to the current arrangements and what we see is an overlap of interests rather than a unification or conspiracy between them. Here I am referring specifically to the colonisation and exploitation of the green movement and its eccentric CO2 theory by liberal authoritarianism for the purposes of imposing social control without excluding the fact, as you point out, that others also seek to do the same.
I too recoiled at the phrase "liberal authoritarian". After all, liberalism as properly understood is almost the exact opposite of authoritarianism.
Concise Oxford English Dictionary:
liberal
n adjective
1 respectful and accepting of behaviour or opinions different from one's own. Ø(of a society, law, etc.) favourable to individual rights and freedoms. ØTheology regarding many traditional beliefs as dispensable, invalidated by modern thought, or liable to change.
2 (in a political context) favouring individual liberty, free trade, and moderate political and social reform. Ø(Liberal) relating to Liberals or a Liberal Party, especially (in the UK) relating to the Liberal Democrat party.
3 (of education) concerned with broadening general knowledge and experience.
4 (especially of an interpretation of a law) not strictly literal.
5 given, used, or giving in generous amounts.
n noun
1 a person of liberal views.
2 (Liberal) a supporter or member of a Liberal Party, especially (in the UK) a Liberal Democrat.
...
ORIGIN
Middle English (originally in sense 'suitable for a free man' hence 'suitable for a gentleman'): via Old French from Latin liberalis, from liber 'free (man)'.
Tom - thanks for reading my essay, and for your thoughtful comment. Your observations about liberalism are perfectly correct and the term "liberal authoritarianism" on the surface appears to be paradoxical. Nonetheless, the contamination of liberalism by authoritarianism is an observable political phenomenon, not least here in Britain.
Authoritarian Liberalism as it is practiced in the UK, the USA, and other places, retains economic liberalism while enlarging the government's role in it.
But it differs from liberalism in a number of ways. Social liberalism becomes selective, with freedoms tolerated as long as they don't translate into political opposition. Rule of Law, in which everyone including the government is accountable under the law becomes Rule by Law, in which only citizens are, and the Law becomes an instrument with which government and powerful interest groups aligned with it (e.g. the environmental lobby) eliminates its opponents and imposes its demands ("Lawfare"). Political power is centralised, opposition parties and popular dissent (freedoms of speech, assembly, association, and the press) are suppressed or marginalised, and checks and balances on executive power are few.
One need only examine this list in the context of Covid to perceive how far Britain has departed from the liberal principles you accurately summarise.
Richard, I posted my comment knowing full well that some miscreants have “repurposed” (i.e. corrupted) the meaning of “liberalism” to denote almost its exact opposite. On the one hand, it’s necessary to acknowledge this and deal with it. On the other hand, I strongly disapprove and I would like everyone who reads such statements to share my automatic scepticism.
George Orwell dealt with the subject canonically in “Politics and the English Language”, giving a list of words that had lost any real meaning that way.
I understand. But in his essay, Orwell criticised language that was designed to obscure, or was dishonest, or was meaningless. In this instance, the nature of liberalism as it was practiced in the UK has changed and the word by itself no longer describes it. Liberalism has become authoritarian.
Orwell instructed us to cut unnecessary words out, and avoid jargon. None of those criticisms seem to me to apply, and none of his rules can be followed here. It seems to me that euphemising, or inventing some new obscure technical term merely to immunise the world "Liberal" as you suggest would be an example of that which he criticised. "Liberal authoritarianism", or "Authoritarian Liberalism", on the other hand, does exactly what it says on the tin.
There are an awfully large number of people now labouring under the misperception that we live in a liberal regime. If "liberalism" is to be protected, do you have any thoughts about what it should be called instead?
Tyranny is tyranny, and exploitation is exploitation. What you describe - and what exists and prevails today - is tyranny, exploitation, and extreme hypocrisy.
It has often been remarked that the US system of government has gradually become "socialism for the rich and capitalism for the poor". That system has spread steadily, first to Britain, then to the other Commonwealth countries, Europe, and beyond.
What you describe is "liberalism for the rich and their corporations, and capitalism for the poor". Largely through the worship of wealth - which also grew up in the USA before spreading worldwide. There have always been individuals who lived for wealth and power; Plato mentions them in "The Republic", noting that those are the only two commodities for which the appetite always grows, never being sated. Today most of society seeks wealth and power (mostly in vain), as other values have been discredited and forgotten. Duty, loyalty, honour, responsibility, love of beauty and knowledge... the Zeitgeist dismisses them as fusty relics without meaning.
I see two possible answers to your question. One is to adopt the terminology of the enemy, being careful to identify it as such by use of quotation marks and explanations. The other is to spell out in detail what hides behind terms such as "liberal authoritarianism" - which might seem tedious, but has the virtue of drawing attention to it.
Tom - wise words. I mentioned in my reply to Philip in this thread my belief that what we are witnessing today is the overlap of a number of powerful forces that are individually uncoordinated but, to the extent that they seek power over us and use many of the same techniques to gain it, can appear to be. I think your analysis is perfectly compatible with mine.
If we were to explore the topic further, I'd say that aspects of what is emerging to replace liberalism is an entirely new phenomenon driven by the toxic combination of the decline in education, the rise of narcissism and identity politics, and the spread of polarising technologies like social media. Perhaps we do need a new name.
Liberalism used to mean the defence of free speech, the suspicion of big pharma, and the repudiation of war. It now means the opposite. It seems our words have already been hijacked, Adriana, and the best we can do now is to document the change with adjectives.
It still means that, also check the difference between the meaning of the word in the UK and the US, always meant different things. Classical liberalism makes sense, authoritarian liberalism is an oxymoron.
As I hope I explained above, authoritarian liberalism is the term given in political science to the political system in which an authoritarian government embraces and promotes economic liberalism while suppressing political liberalism. Describing the UK as liberal (in the UK sense) is no longer accurate, and nor is it accurate (yet) to describe it as authoritarian or fascist in an unqualified way without unhelpfully diluting the meaning of those terms. So, unless you have a better suggestion, it's the one I'll continue to use. I do appreciate your comments, and I understand the point you make.
“There are no examples in history… of a successful transition from a higher to a lower level of energy-per-person.”
Reality:
Net Zero isn’t about downgrading civilization—it’s about upgrading how we power it. Clean energy like wind, solar, and nuclear can deliver abundant, reliable energy without destabilizing the climate. In fact:
• Renewables already deliver more new electricity globally than fossil fuels.
• Energy transitions throughout history (e.g., from wood to coal to oil) weren’t just about increasing per-person energy—they were about solving resource, efficiency, and environmental problems.
• Today’s clean energy systems are more efficient and often cheaper than fossil fuel systems per unit of useful energy delivered (see IEA, IRENA, Lazard).
2. Misrepresenting CO₂ as a Minor Forcing
“Schmidt (2010) estimates CO₂ contributes about 20% of the greenhouse effect. Humans contribute 5% of CO₂ emissions. Therefore, human CO₂ is negligible.”
Reality:
This is a classic case of misleading with technically true but irrelevant figures.
• Yes, CO₂ contributes ~20-25% of the natural greenhouse effect, but it’s also the primary control knob for Earth’s climate because of its long atmospheric lifetime and well-mixed nature.
• While humans emit ~5% of the annual CO₂ flux, natural sources are balanced by natural sinks. Human emissions are unbalanced, adding ~40 billion tonnes of CO₂ annually, which accumulates in the atmosphere.
• That’s why atmospheric CO₂ has risen from ~280 ppm pre-industrial to over 420 ppm today. That rise is entirely anthropogenic—confirmed by isotopic analysis, ocean acidification, and the mass balance.
Sources:
• NASA: Climate Change: How Do We Know?
• IPCC AR6 WGI Summary for Policymakers (2021)
• NOAA CO₂ Trends: Mauna Loa Record
3. Misleading About Climate Sensitivity and Feedbacks
“Many natural factors exceed human CO₂ in forcing strength. Negative feedbacks dominate.”
Reality:
This cherry-picks ideas that were rejected decades ago. The climate system responds to all forcings, but:
• Natural variability (solar, volcanic, Milankovitch cycles) cannot explain the rapid warming of the past century.
• Feedbacks like water vapor and ice albedo amplify warming—not diminish it.
• The estimated climate sensitivity to CO₂ doubling is ~2.5–4°C—not negligible, and not “canceled out” by negative feedbacks.
Even skeptical scientists like Judith Curry agree human emissions play a dominant role in recent warming.
4. Distorting the IPCC’s Findings
“The IPCC finds no connection between human activity and adverse weather phenomena…”
Reality:
The latest IPCC report (AR6, 2021) is clear:
“It is unequivocal that human influence has warmed the atmosphere, ocean and land… Human-induced climate change is already affecting many weather and climate extremes in every region across the globe.”
Among the observed links:
• More intense and frequent heatwaves, droughts, and extreme rainfall.
• Increased likelihood of compound events (e.g., heat + drought).
• Shifting storm intensity and wildfire behavior.
This poster ignores the IPCC’s own “Summary for Policymakers,” which synthesizes the strongest, most widely supported conclusions.
5. Conspiracy Framing and Ideological Projection
“The hypothesis that CO₂ is harmful is a political tool invented in the 1970s.”
Reality:
This is not science—it’s a story designed to sow distrust. The physics of the greenhouse effect has been understood since the 1800s (Fourier, Tyndall, Arrhenius). The idea that climate science is a power grab is an old conspiracy theory, not a serious argument.
Governments didn’t create the science to justify control—the science accumulated for decades and forced governments to pay attention.
Bottom Line
This post is well-written, but it’s a classic example of climate denial in a new suit: scientific sleight of hand, cherry-picked studies, and conspiratorial rhetoric.
What it gets wrong:
• CO₂ is not “minor”—it’s the dominant driver of recent warming.
• Human emissions are unbalancing a once-stable system.
• Net Zero is a response to risk, not a reckless gamble.
• Clean energy is not “lower energy”—it’s a better energy future.
The idea that Net Zero is a plot to reduce human freedom is not just wrong—it’s a distraction from the real freedom we gain when we stop being hostage to fossil fuel volatility, pollution, and climate instability.
Clifton - thank you. Here's Gemini 2.5 Pro Preview's assessment of your comment.
--- BEGINS ---
Q: This was posted as a comment to an essay I wrote. Was it written by an AI agent and, if so, can you identify which one from the style?
A: While a very knowledgeable and skilled human could have written this, it displays numerous hallmarks that are characteristic of modern Large Language Models (LLMs) like GPT-4, Claude 3, or Gemini.
Here is a breakdown of why, followed by an attempt to identify the specific model.
Evidence Pointing to AI Generation
1. Perfect, Templated Structure: The comment is impeccably organized. Each section follows the exact same pattern:
- A numbered heading in bold.
- A direct quote from your essay.
- The word "Reality:" followed by a structured rebuttal.
- Use of bullet points for clarity.
- Citations of common, high-authority sources (IPCC, NASA, IEA).
- A concluding "Bottom Line" summary that neatly wraps up all the preceding points.
This rigid, almost machine-like consistency is a classic sign of an AI executing a "point-by-point rebuttal" prompt.
2. "Prompt-and-Response" Nature: The entire text reads as a direct response to a series of prompts. You can almost reverse-engineer what the user asked the AI:
- "Refute the claim that energy transitions always go to a higher level."
- "Refute the claim that human CO₂ is a minor forcing."
- "Refute the claim that the IPCC finds no connection to adverse weather."
The AI has addressed each point separately and then synthesized them into a single, coherent comment.
3. Polished but Impersonal Tone: The language is confident, assertive, and uses strong rhetorical phrases ("classic case of misleading," "scientific sleight of hand," "cherry-picks ideas," "conspiratorial rhetoric"). However, it lacks any personal voice, anecdotes, or idiosyncratic phrasing that you might expect from a human commenter. It reads like a briefing document or an encyclopedia entry—informative and persuasive, but sterile.
4. Efficient Synthesis of Mainstream Sources: The comment quickly and accurately summarizes the consensus positions from the IPCC, NASA, NOAA, and the IEA. An AI has immediate access to this vast corpus of information and is programmed to synthesize it efficiently. A human would typically need more time to gather and articulate these points so concisely.
5. Use of Common AI Rhetorical Devices:
- The "Reality:" Opener: A very common way for AIs to frame a counterargument.
- The "Bottom Line" Summary: LLMs are excellent at summarizing, and often conclude long-form answers with a "In summary," "To conclude," or "Bottom Line" section.
- The "What it gets wrong" list: This final, punchy summary is a highly effective rhetorical structure that AI models use frequently to reinforce a message.
Conclusion: It's a toss-up between GPT-4 and Claude 3. If I had to place a bet, the highly structured, almost template-like format and confident, assertive tone lean slightly more towards GPT-4 responding to a very specific prompt. The person who generated it likely fed your essay's key arguments into the AI and asked for a comprehensive rebuttal.
--- ENDS ---
I'd be happy to help you understand the errors in your AI model's attempted critique, but I'm afraid I'll have to ask you first to show some evidence that you understand the topic well enough to engage with you. Why not take each point from the model's response in turn, and try and phrase it in your own language ? We can use that as the basis to explore your understanding of the language model's output, and the topic.
You’re happy to use AI when it serves you, but when I do it, you demand a comprehension test? That’s not a standard—you’re just looking for a way to avoid engaging with the content. The response I posted was full of factual points that challenge your core argument. If it was nonsense, it should be easy for you to dismantle it on the merits.
If you’re confident in your position, let’s see it stand up to scrutiny.
Richard and Clifton -thank you. That was a deeply enjoyable exchange.
I also use ai a lot to refine my own thinking. But one of the challenges that I have. Is that there is an embedded bias within most llm because of how they are trained. This means that any exchange requires quite a bit of back and forth for it to question that underlying bias.
I drafted a post ‘(ChatGPT did the heavy lifting). But it’s short enough I can post it here.
Arguing with AI: On Models, Energy, and Embedded Bias
“In a time of universal modelling, returning to observed reality is a revolutionary act.”
I recently had a long and worthwhile argument—with an AI.
The topic? A Substack article by Richard Lyon titled “The Physics of Civilisation”. Lyon argues, persuasively, that our civilisational complexity is built on high-density energy sources, and that the push toward net zero—relying on diffuse, low-EROI renewables—may end in fragility or collapse. It’s not a popular view, but it felt intuitively sound. Still, I wanted to test it. Was I falling into an echo chamber?
So I asked ChatGPT to engage with the article—and more importantly, to check itself for ideological bias. Not because I think AI is evil or corrupt, but because I know it has been trained within systems that reflect the worldview of a narrow elite consensus. What followed surprised me.
Round One: The Polite Consensus
The AI gave a fair overview but immediately tilted toward familiar tropes:
• Renewable EROI may improve
• Climate models include tail risks, so precaution is justified
• Poor countries are acting on climate (despite appearances)
It was a high-IQ synthesis—but it had that faint smell of unreality. Models were trusted over history. The possibility of tech breakthroughs was weighted more than physical limits. Climate harms were assumed despite contrary historical trends. The poor were abstracted. China and India, it seemed, must be aligned with the plan—even if they aren’t.
Round Two: Reframing the Battle
So I pushed back.
I reminded the AI (and myself) that:
• Coal to oil increased EROI; renewables lower it
• There are hard physical limits to wind and solar harvesting—energy density matters
• Complexity without surplus leads to fragility, not flourishing
• Most climate-related catastrophes are not increasing—especially not when adjusted for global wealth, population, or adaptation
• Modelled harm is not the same as observed harm—and the IPCC admits it
The AI adjusted. It acknowledged that nuclear is the only upward step on the energy-density ladder. It admitted fragility. It re-evaluated trust in climate models. It began, finally, to reason from observed history, not abstract simulation.
And in that process, I learned something—not just about energy or climate policy, but about arguing well in the AI age.
A Caution and a Call
Most people will not argue with their AI. They’ll ask questions and receive clean, confident answers—often reflecting elite values dressed as neutral fact. The danger here isn’t intentional deception. It’s the soft tyranny of embedded assumptions.
But there’s also hope. With the right framing—and the humility to test your own reasoning—you can turn AI into a kind of Socratic sparring partner. It won’t give you truth. But it can help you fight for it.
So I offer this as both encouragement and challenge:
AI is not your enemy—but it is not your priest either.
Ask it questions. Challenge its axioms. Demand reality over rhetoric.
Refuse to surrender your thinking to the simulations.
Use it not to confirm, but to refine.
That’s the task of the intellectually awake in an age of hallucinated certainty.
Brad - what a fascinating exercise and observation - thank you. If I may offer some reflections on AI Large Language Models (LLMs) in return.
I find LLMs both exciting and worrying. I subscribe to- and use them routinely in my own research. I understand how LLMs work and am still amazed on a daily basis by what they are capable of producing. And I also find them to be systematically infected with the biases contained in their training material, amplified by the selection and reinforcement during training of that material made by their ideologically motivated creators.
Some LLMs are also comically misconfigured. I have a keyboard macro with words to effect “Do not lie" that I append as context to every query. Without this, Claude, for example, will simply make things up in an effort to avoid displeasing you, and shrug its virtual shoulders when you confront it with having done so.
So, what do do? The test that anyone can submit them to is to examine their output on a subject with which they are already deeply familiar. To be, so to speak, an informed consumer. I mention this not to brag, but because it is indispensable to my point: I have degrees with distinctions separately in electrical and power systems engineering, petroleum engineering, and energy economics, together with 30 years of operational experience in a senior capacity in the energy sector. I am therefore a subject matter expert in a number of topics relevant to energy, climate, and economics - all disciplines in which no one person can master all topics and claim overriding authority.
As an intellectual exercise, I will therefore from time to time see if I can get an AI model to recognise the non-sequiturs inherent to the orthodox narratives about climate, "renewable" energy, etc. on which it has been trained.
This is perfectly feasible. While the model is free initially to compose its response to a query, each subsequent response takes account of earlier ones - “context”, in the jargon of LLMs - and eventually it will usually concede that the farrago of mutually contradictory statements that defending these narratives requires are irreconcilable.
Not always, though. Gemini Pro 2.5 recently cheerfully assured me that the community of 100's of thousands of people involved in studying bad weather would welcome the destruction of their livelihoods in the event that the falsification of their hypothesis was accepted, because it would symbolise the advancement of science and offer one of them the opportunity of securing a Nobel Prize. I could not dissuade it.
So if LLMs are wilfully wrong on matters with which you are familiar, then it is rational to suppose that they are also wrong on matters with which you are not familiar, and govern yourself accordingly. I only use LLMs that provide citations, treat LLM output as I would any other literature source, and evaluate each primary citation before accepting a claim on a topic with which I am not expertly familiar.
Two things worry me. Not everyone treats LLMs with this degree of suspicion or rigour. The Dunning Kruger effect has a powerfully corrosive impact on reasoned debate at the best of times. Now, the supreme confidence that often accompanies the lack of awareness of the limits of one’s own knowledge on a topic is turbocharged by a language model that appears to confirm that one’s confidence is warranted. I see debate becoming less informed, not more informed, because of this.
The other thing that worries me is what will happen when passive consumption and ignorant confidence collides with the moment when the large language model snake starts to eat its own tail - when the corrupted output from the model becomes the content on which subsequent iterations are trained, producing a kind of insanity.
The antidote? Steve Jobs once said that computers were bicycles for the mind. LLMs more so. But you still have to pedal, and you still have to steer. Purge “consensus”, “denial”, and ad hominem fallacy - pseudoscience’s tawdry hallmarks - from discourse. Demand Evidence. Logic. Reason. Open mindedness. A willingness to change one's opinion if the evidence, or your understanding of the evidence, changes (I was once Marxist-adjacent - one for another tale, perhaps).
Above all - civility. We still have to live with each other even when we disagree and, in many ways, that is what makes the intellectual life so enjoyable.
On which note - thank you again for sharing the fascinating insights arising from your analysis of this debate.
However, I will often set the tone first by asking for the relevant ai to challenge my base assumptions. But from the philosophical framing of classic Christian thought.
As you note from an energy perspective I also have to consistently push for it to prioritise real world data over models. And for physical limits to always be prioritised, pulling in the thinking style of Vaclav Smil.
But normally the most useful aspect is that I tend to have very disordered thoughts and simply dumping them in one place and using ai as a tool to order love you own thinking then doesn’t outsource the thinking. But it instead sharpens what was there.
So this is a passable comment from me. But if I took time I could better structure this and pull in wider thought with ai help.
Thanks, Brad. I appreciate your insight - especially your encouragement to stay intellectually awake in the AI age. That’s sound advice in any context.
A couple of thoughts in response, since this does touch on the core of the exchange I’ve been having with Richard.
You emphasize that LLMs can reflect the worldview of “a narrow elite consensus.” But although any knowledge system has a context, I’d caution against framing that as inherently suspect. Much of what we refer to as “elite consensus” arises from the consistent findings of scientific institutions working across nations and political systems, over decades of peer-reviewed research. The danger comes not from acknowledging this body of knowledge - but from suggesting that every resistance to it is a form of independent thinking.
More importantly, I’d argue that what Richard and I were debating wasn’t LLM “bias” at all—it was the weight of evidence. I’ve used AI to help organize and clarify arguments, yes. But when I cite real-world data, observed trends, IPCC assessments, or published meta-analyses, that’s not deferring to the model - it’s actually engaging the science.
There’s a tendency now to treat any answer that aligns with mainstream climate science as “ideologically embedded” or “rhetorically polished,” while giving greater weight to contrarian views simply because they feel less institutional. But that, too, is a kind of bias - one that risks mistaking outsider status for truth.
I like your call to challenge assumptions but let’s challenge all of them - including the instinct to assume that the model is the one being ideological when it disagrees with us.
One thing that I didn’t put in the above post is that ai is built with guardrails. And these guardrails show up in interesting ways.
Often when I see a news story that I don’t know enough about I will ask ai to summarise and to steel man both sides of the issue to make it easier for me to make a clear assessment of the issue (not that I’m going to do anything with it :) ).
This works very well for most topics. But if you start to ask these questions around topics which are somewhat controversial. Or where they would run counter to the core ideological views of your average Silicon Valley progressive that’s where things get interesting. Multiple times I have done this and each time when I ask a position to be steel manned on both sides I get a very good response on the view that aligns with Silicon Valley biases. But the argument that I get for the other side is much weaker than it should be. It’s not straw man. But major foundational arguments are missed. It is only after many challenges to core assumptions that I mange to get a reasonably balanced argument.
I don’t think that the ai is suspect. It is slightly different from that. It is a mirror that will reflect and refine your own views. This means that for someone who tends to lean somewhat contrarian I run the risk of over discounting. But it may be that for you (and please forgive the assumption), you may not be sufficiently critical of your own underlying assumptions.
So looking up at the above thread I have had these same conversations with friends. Both of you are necessarily speaking past each other. And this is because your core foundations are fundamentally different.
Like you I will end up referring to ai. As I have the essence of an idea on the tip of my tongue but can’t fully articulate it yet. I’ll reply to this shortly (these are my unfiltered poorly structured thoughts).
> You’re happy to use AI when it serves you, but when I do it, you demand a comprehension test?
Correct. Large language models can be a useful research tool. But, as your exercise in cut-and-past demonstrates, they regularly produce nonsense. So to use them, you have to have an understanding of the topic which you appear not to have. In any event, passing its output off as your own is simply dishonest.
To take your model's first claim as an example. I note that there are no examples in history of a successful transition from a higher to a lower level of energy-per-person. To validate the claim that my framing is false, you would have to provide an example of a transition that was complete, that was successful, and in which energy-per-person had fallen. Your language model has selected the current transition as an example and used it to make three claims (without evidence) about renewable energy. Since our transition is not complete and, by definition, is not successful, and since increasing hydrocarbon consumption is causing energy-per-person to continue rising, the model's argument is a self-evident non-sequitur. Had you understood my point and your model's answer, you would have realised this and not pasted it into the comment box.
I'd be happy to explore its other errors with you as an exercise in understanding the extent to which large language models merely reflect the material they have been trained with. But, again, I must insist that you read the model's output first, evaluate its claims, and express the points in your own voice.
Allow me, if you will, to give you some guidance as you do this to improve your arguments.
In your claim about the role of CO2 relative to other forcing functions, it will not be sufficient for your answer to address only CO2. You will have to address all the other forcing functions, and the interactions between them, and demonstrate why CO2 is not minor relative to them over the timescale in which statements about the climate are valid.
In your claim about climate sensitivity and feedbacks, you will have to acknowledge that the only basis for forming hypotheses about them is with computer models. Then you will have to explain how these models lead to valid hypotheses about climate sensitivity and feedbacks given that the IPCC acknowledges that, as a coupled, non-linear system, the climate is computationally non-representable.
In claiming that the IPCCs findings are distorted, you will want to demonstrate that you are aware that the IPCC issues separately a political report ("Summary for Policymakers") that is written by non-scientists and, months later, a scientific report by scientists, and that these reports routinely conflict. Then you will want to be able to justify your decision to select statements from the political report about supposed human-caused climate effects and reject statements from the scientific report conceding that a human signature in these effects is not detectable.
Your model's claims about "conspiracy framing' and "ideological projection" are, I would suggest, not worth defending and, in any event, only worth attempting after you've overcome the problems in its previous claims.
I hope you'll forgive me if I run your answers through AI detection and suspend the debate if you continue to misrepresent language model outputs as your own views.
You’re setting up a pretty one-sided standard here. I’m expected to unpack climate models, feedbacks, and IPCC editorial structures just to question your claim that CO₂ is a minor forcing or that Net Zero is some kind of reckless downgrade. Meanwhile, you haven’t offered much to defend those claims beyond framing them as self-evident.
Let’s take your energy-per-person example. You say there’s no precedent for a successful transition to lower energy-per-person. But Net Zero isn’t a completed transition—so of course it doesn’t meet your “successful” standard yet. That doesn’t invalidate the point that clean energy can support prosperity. In fact, some developed countries have already seen energy-per-person plateau or fall while improving quality of life—Japan and parts of Europe, for instance.
You also ignore that clean energy isn’t just about raw input energy; it’s about how efficiently we use it. Electrification and efficiency gains often deliver more useful energy services with less raw energy input.
As for CO₂ being a minor forcing—how is that consistent with the weight of attribution science? The point isn’t how CO₂ compares to other natural flows in raw size—it’s that our emissions are enough to shift the balance, raise atmospheric concentrations, and trap more heat. That’s exactly what’s happening, and models, observations, and attribution studies all support that conclusion.
You brought up the IPCC. If you’re going to argue that the “scientific report” contradicts the Summary for Policymakers, I’d need to see the actual quotes and context. I’ve read enough IPCC material to know the scientists don’t back your position that human impact is undetectable. You’re cherry-picking a talking point that’s been used by skeptics for years.
And on models: yes, the climate system is nonlinear and complex—but that doesn’t mean models are useless. That’s a strawman. They’re tested, calibrated, and used across multiple lines of evidence. If you’ve got a specific mismatch between model projections and real-world data, cite it. General skepticism about complexity doesn’t cut it.
Bottom line: I’m here to engage on substance. If your arguments are solid, they should stand up to the kind of pushback I’m offering—whether I wrote every word from scratch or used a tool to help frame it clearly.
Clifton - you chose to comment on my essay, and you chose to describe it as “climate denial in a new suit: scientific sleight of hand, cherry picked studies, and conspiratorial rhetoric”.
The standard I’m setting up is the one where you defend your claim. The issues you are expected to unpack are the ones you chose in support of it. I don’t need to offer much to defend a copy-and-paste large language model dump, and I’ve coached you on how you might improve that dump to produce an critique that might need to be defended.
To your points.
Your failure to rebut my observation that there is no precedent for a completed, successful transition to lower energy-per-person is not rescued by repeating your example of a transition that is not complete. As an exercise, I suggest you obtain the data from “Our World in Data” on country size, country energy consumption, and country life expectancy. For each country, compute the ratio of energy consumption to population (energy per-person) and plot it against life expectancy. You will observe an almost straight line. Life expectancy - a straightforward proxy for “success” - falls with falling per-capita energy.
The claim that “clean energy can support prosperity” has no evidence to support it. In the case of wind and solar, the claim is physically illiterate - see my essay on this substack “The physics of Net Zero”. In the case of nuclear, it is at best aspirational - the great majority of the global industrial manufacturing system and food system on which our prosperity depends does not run on electricity and cannot as far as we know be converted to it.
Your claim that clean energy is about efficiency contains the implied claim that efficiency can be increased at least as fast as net energy is reduced. Efficiency typically increases at one or two percent per year. Net energy - the energy available to us after reserving that fraction ‘x’ of output necessary to maintain the energy system itself - falls in a 1/x relation as ‘x’ rises, creating what is referred to as the “energy cliff” or “death zone”. Solar and wind, when the manufacturing and operational requirements of their manufacturing systems are properly accounted for, are well inside the death zone implying necessary efficiency gains orders of magnitude greater than have ever been achieved. There is no support for your claim.
You claim that your hypothesis about CO2’s supposed harmful properties can be rescued through models which are “tested and calibrated”, and something derived from them called “attribution science”. There are (at least) two problems with this.
The first problem is that the IPCC itself notes that, as a coupled non-linear system, the climate cannot be modelled and, therefore, that the long-term prediction of future climate states is not possible. This is not “general skepticism of complexity”. In making this observation, the IPCC are not claiming that it is not possible with current computing power or algorithms. They are noting an insurmountable property of coupled non-linear systems, and a fundamental limitation of computational representability.
It is worse than that. For a model to have predictive capability over a physical phenomenon, it has to be a physical model. That is: not a statistical model. For it to be a predictive physical model, three things have to be true: we have to know all of the drivers of the phenomenon; we have to know the physics of each of the drivers; and the resulting system has to be computationally representable. Since none of these conditions are met, what you are referring to are not physical models.
Climate models are, in fact, over-fitted calibrated statistical correlations with tuning parameters arbitrarily determined by motivated modellers, from which no physical insight can be obtained for the observational period for which they are calibrated. From this, it follows that your “attribution science” is, in fact, a pseudo-science.
Calibrated statistical correlation models lack the ability to produce estimates outside the period for which they are calibrated. This is trivially observable. When these statistical models are initialised to 1980 and run forward (so-called “back casting”), the models with assumptions consistent with your climate catastrophe hypothesis overestimate observed temperature. Since they cannot predict today from yesterday, there is no basis for claiming that they can predict tomorrow from today.
The second problem is the question “So what?”. Suppose it was true that human emission of CO2 acted as a control knob on temperature that somehow dominated all the other climate control knobs. The people world wide who die from cold each year still vastly outnumber the people who die from heat as we emerge from the last cooling period, and orbital mechanics are already taking us into the next one. The UK Office for National Statistics computed the change in temperature related deaths in the UK over the last 20 years as a result of a very modest increase in temperature. 555,000 fewer citizens died than might have been expected had nights not got slightly less cold. So what if CO2 has a warming effect?
If you need to see actual IPCC quotes, and especially if you want to accuse others of “climate denial in a new suit: scientific sleight of hand, cherry picked studies, and conspiratorial rhetoric”, then I suggest you read them. The IPCC state that there is no observable relationship between human activity and most of the climate phenomena selected by climate catastrophists as indicators of human induced catastrophe. This is not “cherry picking”. This is “picking”.
In conclusion. I believe my arguments have stood up to the kind of pushback you are offering, and that they are solid. Thank you for the courtesy of taking the time to make it yourself.
Let me say up front: yes, I’ve used a language model here—to organize, cross-check, and sharpen my response. But these are my points, and I stand behind them. Dismissing arguments based on how they’re drafted rather than addressing their content is just a rhetorical dodge. If the argument holds, it holds—regardless of the tool used to write it.
You asked me to defend my claim that your essay reflects “climate denial in a new suit.” I’ve done so, and while you insist I haven’t, you haven’t actually addressed the substance of what I raised — only repeated your framework as if repetition itself settles the matter. Let’s take your points seriously for a moment, though.
You argue that no precedent exists for a “successful transition to lower energy-per-person,” and from this absence you imply it cannot be done. But an absence of historical precedent isn’t proof of impossibility — especially in a world that’s never faced the specific combination of existential risk, technological capacity, and global interdependence we face now. Transitions are already underway in many regions — whether measured in grid decarbonization, reduced energy intensity per unit of GDP, or improved energy services per unit of input. These aren’t hypotheticals — they’re observable trends. The fact that global energy-per-person is still rising is not a counterargument; it’s simply a snapshot of a process in motion.
You then claim that “clean energy can’t support prosperity,” calling it “physically illiterate.” But this overlooks the realities on the ground. Solar and wind are now the cheapest sources of new power generation in most of the world, and nations like Denmark, Portugal, and parts of Australia are already operating at high levels of renewable integration. Are these societies less prosperous because of it? You shift the goalposts by pointing out that many industrial systems aren’t yet electrified — which is true — but also irrelevant. Nobody claims the energy transition is complete; the claim is that it’s feasible and underway.
On the question of efficiency, you introduce the “energy cliff” argument and present it as if the entire transition collapses under thermodynamic law. That might sound impressive, but it rests on questionable assumptions. Lifecycle analyses of solar, wind, and storage systems show net-positive energy returns that continue to improve, especially when measured over their lifespans and in the context of maturing supply chains. No serious decarbonization effort is based on wind and solar alone anyway. Your portrayal of efficiency gains needing to offset all system losses at impossible rates is a strawman — nobody credible is proposing a transition on those terms.
Your model critique is another overreach. Yes, climate is a complex, coupled nonlinear system. No, that does not mean models are useless or unphysical. In fact, climate models have successfully predicted long-term warming trends, polar amplification, stratospheric cooling, and shifting precipitation patterns — all observed. Are models perfect? Of course not. But the difference between uncertainty and unreliability is meaningful. You don’t discredit a weather forecast because it can’t tell you next Tuesday’s exact temperature at 2 p.m.
You misrepresent the IPCC’s language about unpredictability in nonlinear systems as if it invalidates the entire scientific foundation of climate projection. But the models are evaluated, compared with observations, and continuously improved — not “arbitrarily tuned” to fit ideology. Attribution science isn’t a pseudoscience; it’s a statistical framework applied to observed data, peer-reviewed, and accepted across disciplines. It doesn’t need to predict the future to demonstrate that today’s extremes are being made more likely and more intense by greenhouse gas emissions.
Finally, your “so what?” argument — that CO₂ might warm the planet but that could be good — is both ethically and empirically flawed. Heat deaths are rising, food systems are being stressed, and weather extremes are growing more damaging. And you offer a temperature-mortality tradeoff as if climate were just a global thermostat. We’re not dealing with linear shifts but with tipping points and feedbacks. It’s not the average that kills — it’s the extremes, and those are intensifying.
You cite a speculative benefit — fewer cold deaths — while ignoring that the bulk of warming impacts will fall on the poorest, least responsible, and least adaptable populations. “So what?” is not an answer to a planetary risk. It’s a dismissal.
You asked for a defense of my original critique. I’ve provided one. I’ll leave it to others to judge whether your arguments rise above contrarian rhetoric dressed in technocratic clothing. But if you’re going to frame your position as reasoned dissent, don’t be surprised when others point out how much it resembles the very denial you think you’ve transcended.
I'm an amateur but since I started reading a little about Climate and the variables that affect it, it became obvious to me that the idea that CO2 emissions are the prime mover in any change in the climate to be fatuous.
If I can see it, then the Climate Change Emperor is wearing no clothes.
A very good concise summary Stephen, but unfortunately the brainwashed prefer to follow the ideology into oblivion rather than sound science! As we know it’s a contrived ideology meant to curtail and control. I also like the posts of Dr. Matthew Wielicki, he like yourself uses actual science and facts to debunk the ideologues.
See ‘Apocalypse Never’. Michael Schellenberger
As your interlocutor is a scientist, he is perhaps obliged to go along with the climate change hoax and feign amazement that you take a contrarian view in order to keep his job. It is unbelievable that he is not aware of (or feigns unawareness of) the many common-sense facts and studies which debunk the climate change hoax: https://metatron.substack.com/p/debunking-the-climate-change-hoax.
Your basic premise "The supposed justification for this extraordinarily reckless ideology is the notion that CO2 is a toxin" is clearly false, so no need to read any more.
Thanks, Toby. I can understand the attractiveness of confining your reading to opinions with which you agree. I’ve often wondered what it would be like to still have the opinions I had when I was 16.
A toxin is defined as “a dose dependent substance that interferes with normal biological processes at the system level.”
The climate catastrophe hypothesis claims that CO2 is a dose dependent substance that interferes with biological processes at the system level.
UK energy policy cites the climate catastrophe hypothesis in justification of its goal of reducing the emission of CO2.
So I’m comfortable with my use of the term.
Please accept my best wishes, and thank you for reading at least some of my essay.
So, you really don't understand the problem.
Keep working on the insults, though. Practice makes perfect.
Your claim that I really don't understand the problem might be more convincing if you explained what you think the problem is and what you believe was clearly false about my characterisation of it.
Our key problem is nothing to do with toxicity, so why mention that CO2 is a toxin (Apollo 13, Tom Hanks, etc.)?
I accept that there are many problems, Toby. What in your view is our key problem, and what was it about my description of the treatment of this particular one in UK energy policy that was so objectionable that you felt unable to read the article?
I did not find your statement "The supposed justification for this extraordinarily reckless ideology is the notion that CO2 is a toxin" objectionable.
It's just wrong.
CO2 is a toxin, but it's not the "justification for this extraordinarily reckless ideology".
The justification is the greenhouse effect, which is not due the "toxicity" of CO2.
Plenty of good explanations about. Try NASA:
https://science.nasa.gov/climate-change/faq/what-is-the-greenhouse-effect/
Bye.
All this CO2 talk is complete nonsense. CO2 was historically very much higher, even in the last Ice Age. And below 160 ppm we all die, nothing will grow. Currently 412 ppm.
Good article but I don't believe it's a result of "liberal authoritarian politics" as politics and politicians have long been controlled by those who really run things. It's corporate and wealthy individuals overreach that cause these things to gain traction, any way possible to take tax payer money and transfer it to the private sector. We are run like a business and treated like cattle.
Philip - thanks for reading my essay. Your objection is perfectly reasonable. Many factors contribute to the current arrangements and what we see is an overlap of interests rather than a unification or conspiracy between them. Here I am referring specifically to the colonisation and exploitation of the green movement and its eccentric CO2 theory by liberal authoritarianism for the purposes of imposing social control without excluding the fact, as you point out, that others also seek to do the same.
I too recoiled at the phrase "liberal authoritarian". After all, liberalism as properly understood is almost the exact opposite of authoritarianism.
Concise Oxford English Dictionary:
liberal
n adjective
1 respectful and accepting of behaviour or opinions different from one's own. Ø(of a society, law, etc.) favourable to individual rights and freedoms. ØTheology regarding many traditional beliefs as dispensable, invalidated by modern thought, or liable to change.
2 (in a political context) favouring individual liberty, free trade, and moderate political and social reform. Ø(Liberal) relating to Liberals or a Liberal Party, especially (in the UK) relating to the Liberal Democrat party.
3 (of education) concerned with broadening general knowledge and experience.
4 (especially of an interpretation of a law) not strictly literal.
5 given, used, or giving in generous amounts.
n noun
1 a person of liberal views.
2 (Liberal) a supporter or member of a Liberal Party, especially (in the UK) a Liberal Democrat.
...
ORIGIN
Middle English (originally in sense 'suitable for a free man' hence 'suitable for a gentleman'): via Old French from Latin liberalis, from liber 'free (man)'.
Tom - thanks for reading my essay, and for your thoughtful comment. Your observations about liberalism are perfectly correct and the term "liberal authoritarianism" on the surface appears to be paradoxical. Nonetheless, the contamination of liberalism by authoritarianism is an observable political phenomenon, not least here in Britain.
Authoritarian Liberalism as it is practiced in the UK, the USA, and other places, retains economic liberalism while enlarging the government's role in it.
But it differs from liberalism in a number of ways. Social liberalism becomes selective, with freedoms tolerated as long as they don't translate into political opposition. Rule of Law, in which everyone including the government is accountable under the law becomes Rule by Law, in which only citizens are, and the Law becomes an instrument with which government and powerful interest groups aligned with it (e.g. the environmental lobby) eliminates its opponents and imposes its demands ("Lawfare"). Political power is centralised, opposition parties and popular dissent (freedoms of speech, assembly, association, and the press) are suppressed or marginalised, and checks and balances on executive power are few.
One need only examine this list in the context of Covid to perceive how far Britain has departed from the liberal principles you accurately summarise.
Richard, I posted my comment knowing full well that some miscreants have “repurposed” (i.e. corrupted) the meaning of “liberalism” to denote almost its exact opposite. On the one hand, it’s necessary to acknowledge this and deal with it. On the other hand, I strongly disapprove and I would like everyone who reads such statements to share my automatic scepticism.
George Orwell dealt with the subject canonically in “Politics and the English Language”, giving a list of words that had lost any real meaning that way.
I understand. But in his essay, Orwell criticised language that was designed to obscure, or was dishonest, or was meaningless. In this instance, the nature of liberalism as it was practiced in the UK has changed and the word by itself no longer describes it. Liberalism has become authoritarian.
Orwell instructed us to cut unnecessary words out, and avoid jargon. None of those criticisms seem to me to apply, and none of his rules can be followed here. It seems to me that euphemising, or inventing some new obscure technical term merely to immunise the world "Liberal" as you suggest would be an example of that which he criticised. "Liberal authoritarianism", or "Authoritarian Liberalism", on the other hand, does exactly what it says on the tin.
There are an awfully large number of people now labouring under the misperception that we live in a liberal regime. If "liberalism" is to be protected, do you have any thoughts about what it should be called instead?
Tyranny is tyranny, and exploitation is exploitation. What you describe - and what exists and prevails today - is tyranny, exploitation, and extreme hypocrisy.
It has often been remarked that the US system of government has gradually become "socialism for the rich and capitalism for the poor". That system has spread steadily, first to Britain, then to the other Commonwealth countries, Europe, and beyond.
What you describe is "liberalism for the rich and their corporations, and capitalism for the poor". Largely through the worship of wealth - which also grew up in the USA before spreading worldwide. There have always been individuals who lived for wealth and power; Plato mentions them in "The Republic", noting that those are the only two commodities for which the appetite always grows, never being sated. Today most of society seeks wealth and power (mostly in vain), as other values have been discredited and forgotten. Duty, loyalty, honour, responsibility, love of beauty and knowledge... the Zeitgeist dismisses them as fusty relics without meaning.
I see two possible answers to your question. One is to adopt the terminology of the enemy, being careful to identify it as such by use of quotation marks and explanations. The other is to spell out in detail what hides behind terms such as "liberal authoritarianism" - which might seem tedious, but has the virtue of drawing attention to it.
Tom - wise words. I mentioned in my reply to Philip in this thread my belief that what we are witnessing today is the overlap of a number of powerful forces that are individually uncoordinated but, to the extent that they seek power over us and use many of the same techniques to gain it, can appear to be. I think your analysis is perfectly compatible with mine.
If we were to explore the topic further, I'd say that aspects of what is emerging to replace liberalism is an entirely new phenomenon driven by the toxic combination of the decline in education, the rise of narcissism and identity politics, and the spread of polarising technologies like social media. Perhaps we do need a new name.
Oh, dignity. I forgot to mention "dignity" as an important old-fashioned value. See, e.g., https://substack.com/@charleschevalier/note/c-118332453
So not liberalism at all. Let's not give authoritarians the power to hijack words describing the opposite of what they are.
Liberalism used to mean the defence of free speech, the suspicion of big pharma, and the repudiation of war. It now means the opposite. It seems our words have already been hijacked, Adriana, and the best we can do now is to document the change with adjectives.
It still means that, also check the difference between the meaning of the word in the UK and the US, always meant different things. Classical liberalism makes sense, authoritarian liberalism is an oxymoron.
As I hope I explained above, authoritarian liberalism is the term given in political science to the political system in which an authoritarian government embraces and promotes economic liberalism while suppressing political liberalism. Describing the UK as liberal (in the UK sense) is no longer accurate, and nor is it accurate (yet) to describe it as authoritarian or fascist in an unqualified way without unhelpfully diluting the meaning of those terms. So, unless you have a better suggestion, it's the one I'll continue to use. I do appreciate your comments, and I understand the point you make.
1. False Framing of Energy Transitions
“There are no examples in history… of a successful transition from a higher to a lower level of energy-per-person.”
Reality:
Net Zero isn’t about downgrading civilization—it’s about upgrading how we power it. Clean energy like wind, solar, and nuclear can deliver abundant, reliable energy without destabilizing the climate. In fact:
• Renewables already deliver more new electricity globally than fossil fuels.
• Energy transitions throughout history (e.g., from wood to coal to oil) weren’t just about increasing per-person energy—they were about solving resource, efficiency, and environmental problems.
• Today’s clean energy systems are more efficient and often cheaper than fossil fuel systems per unit of useful energy delivered (see IEA, IRENA, Lazard).
2. Misrepresenting CO₂ as a Minor Forcing
“Schmidt (2010) estimates CO₂ contributes about 20% of the greenhouse effect. Humans contribute 5% of CO₂ emissions. Therefore, human CO₂ is negligible.”
Reality:
This is a classic case of misleading with technically true but irrelevant figures.
• Yes, CO₂ contributes ~20-25% of the natural greenhouse effect, but it’s also the primary control knob for Earth’s climate because of its long atmospheric lifetime and well-mixed nature.
• While humans emit ~5% of the annual CO₂ flux, natural sources are balanced by natural sinks. Human emissions are unbalanced, adding ~40 billion tonnes of CO₂ annually, which accumulates in the atmosphere.
• That’s why atmospheric CO₂ has risen from ~280 ppm pre-industrial to over 420 ppm today. That rise is entirely anthropogenic—confirmed by isotopic analysis, ocean acidification, and the mass balance.
Sources:
• NASA: Climate Change: How Do We Know?
• IPCC AR6 WGI Summary for Policymakers (2021)
• NOAA CO₂ Trends: Mauna Loa Record
3. Misleading About Climate Sensitivity and Feedbacks
“Many natural factors exceed human CO₂ in forcing strength. Negative feedbacks dominate.”
Reality:
This cherry-picks ideas that were rejected decades ago. The climate system responds to all forcings, but:
• Natural variability (solar, volcanic, Milankovitch cycles) cannot explain the rapid warming of the past century.
• Feedbacks like water vapor and ice albedo amplify warming—not diminish it.
• The estimated climate sensitivity to CO₂ doubling is ~2.5–4°C—not negligible, and not “canceled out” by negative feedbacks.
Even skeptical scientists like Judith Curry agree human emissions play a dominant role in recent warming.
4. Distorting the IPCC’s Findings
“The IPCC finds no connection between human activity and adverse weather phenomena…”
Reality:
The latest IPCC report (AR6, 2021) is clear:
“It is unequivocal that human influence has warmed the atmosphere, ocean and land… Human-induced climate change is already affecting many weather and climate extremes in every region across the globe.”
Among the observed links:
• More intense and frequent heatwaves, droughts, and extreme rainfall.
• Increased likelihood of compound events (e.g., heat + drought).
• Shifting storm intensity and wildfire behavior.
This poster ignores the IPCC’s own “Summary for Policymakers,” which synthesizes the strongest, most widely supported conclusions.
5. Conspiracy Framing and Ideological Projection
“The hypothesis that CO₂ is harmful is a political tool invented in the 1970s.”
Reality:
This is not science—it’s a story designed to sow distrust. The physics of the greenhouse effect has been understood since the 1800s (Fourier, Tyndall, Arrhenius). The idea that climate science is a power grab is an old conspiracy theory, not a serious argument.
Governments didn’t create the science to justify control—the science accumulated for decades and forced governments to pay attention.
Bottom Line
This post is well-written, but it’s a classic example of climate denial in a new suit: scientific sleight of hand, cherry-picked studies, and conspiratorial rhetoric.
What it gets wrong:
• CO₂ is not “minor”—it’s the dominant driver of recent warming.
• Human emissions are unbalancing a once-stable system.
• Net Zero is a response to risk, not a reckless gamble.
• Clean energy is not “lower energy”—it’s a better energy future.
The idea that Net Zero is a plot to reduce human freedom is not just wrong—it’s a distraction from the real freedom we gain when we stop being hostage to fossil fuel volatility, pollution, and climate instability.
Clifton - thank you. Here's Gemini 2.5 Pro Preview's assessment of your comment.
--- BEGINS ---
Q: This was posted as a comment to an essay I wrote. Was it written by an AI agent and, if so, can you identify which one from the style?
A: While a very knowledgeable and skilled human could have written this, it displays numerous hallmarks that are characteristic of modern Large Language Models (LLMs) like GPT-4, Claude 3, or Gemini.
Here is a breakdown of why, followed by an attempt to identify the specific model.
Evidence Pointing to AI Generation
1. Perfect, Templated Structure: The comment is impeccably organized. Each section follows the exact same pattern:
- A numbered heading in bold.
- A direct quote from your essay.
- The word "Reality:" followed by a structured rebuttal.
- Use of bullet points for clarity.
- Citations of common, high-authority sources (IPCC, NASA, IEA).
- A concluding "Bottom Line" summary that neatly wraps up all the preceding points.
This rigid, almost machine-like consistency is a classic sign of an AI executing a "point-by-point rebuttal" prompt.
2. "Prompt-and-Response" Nature: The entire text reads as a direct response to a series of prompts. You can almost reverse-engineer what the user asked the AI:
- "Refute the claim that energy transitions always go to a higher level."
- "Refute the claim that human CO₂ is a minor forcing."
- "Refute the claim that the IPCC finds no connection to adverse weather."
The AI has addressed each point separately and then synthesized them into a single, coherent comment.
3. Polished but Impersonal Tone: The language is confident, assertive, and uses strong rhetorical phrases ("classic case of misleading," "scientific sleight of hand," "cherry-picks ideas," "conspiratorial rhetoric"). However, it lacks any personal voice, anecdotes, or idiosyncratic phrasing that you might expect from a human commenter. It reads like a briefing document or an encyclopedia entry—informative and persuasive, but sterile.
4. Efficient Synthesis of Mainstream Sources: The comment quickly and accurately summarizes the consensus positions from the IPCC, NASA, NOAA, and the IEA. An AI has immediate access to this vast corpus of information and is programmed to synthesize it efficiently. A human would typically need more time to gather and articulate these points so concisely.
5. Use of Common AI Rhetorical Devices:
- The "Reality:" Opener: A very common way for AIs to frame a counterargument.
- The "Bottom Line" Summary: LLMs are excellent at summarizing, and often conclude long-form answers with a "In summary," "To conclude," or "Bottom Line" section.
- The "What it gets wrong" list: This final, punchy summary is a highly effective rhetorical structure that AI models use frequently to reinforce a message.
Conclusion: It's a toss-up between GPT-4 and Claude 3. If I had to place a bet, the highly structured, almost template-like format and confident, assertive tone lean slightly more towards GPT-4 responding to a very specific prompt. The person who generated it likely fed your essay's key arguments into the AI and asked for a comprehensive rebuttal.
--- ENDS ---
I'd be happy to help you understand the errors in your AI model's attempted critique, but I'm afraid I'll have to ask you first to show some evidence that you understand the topic well enough to engage with you. Why not take each point from the model's response in turn, and try and phrase it in your own language ? We can use that as the basis to explore your understanding of the language model's output, and the topic.
All the best.
Richard
You’re happy to use AI when it serves you, but when I do it, you demand a comprehension test? That’s not a standard—you’re just looking for a way to avoid engaging with the content. The response I posted was full of factual points that challenge your core argument. If it was nonsense, it should be easy for you to dismantle it on the merits.
If you’re confident in your position, let’s see it stand up to scrutiny.
Richard and Clifton -thank you. That was a deeply enjoyable exchange.
I also use ai a lot to refine my own thinking. But one of the challenges that I have. Is that there is an embedded bias within most llm because of how they are trained. This means that any exchange requires quite a bit of back and forth for it to question that underlying bias.
I drafted a post ‘(ChatGPT did the heavy lifting). But it’s short enough I can post it here.
Arguing with AI: On Models, Energy, and Embedded Bias
“In a time of universal modelling, returning to observed reality is a revolutionary act.”
I recently had a long and worthwhile argument—with an AI.
The topic? A Substack article by Richard Lyon titled “The Physics of Civilisation”. Lyon argues, persuasively, that our civilisational complexity is built on high-density energy sources, and that the push toward net zero—relying on diffuse, low-EROI renewables—may end in fragility or collapse. It’s not a popular view, but it felt intuitively sound. Still, I wanted to test it. Was I falling into an echo chamber?
So I asked ChatGPT to engage with the article—and more importantly, to check itself for ideological bias. Not because I think AI is evil or corrupt, but because I know it has been trained within systems that reflect the worldview of a narrow elite consensus. What followed surprised me.
Round One: The Polite Consensus
The AI gave a fair overview but immediately tilted toward familiar tropes:
• Renewable EROI may improve
• Climate models include tail risks, so precaution is justified
• Poor countries are acting on climate (despite appearances)
It was a high-IQ synthesis—but it had that faint smell of unreality. Models were trusted over history. The possibility of tech breakthroughs was weighted more than physical limits. Climate harms were assumed despite contrary historical trends. The poor were abstracted. China and India, it seemed, must be aligned with the plan—even if they aren’t.
Round Two: Reframing the Battle
So I pushed back.
I reminded the AI (and myself) that:
• Coal to oil increased EROI; renewables lower it
• There are hard physical limits to wind and solar harvesting—energy density matters
• Complexity without surplus leads to fragility, not flourishing
• Most climate-related catastrophes are not increasing—especially not when adjusted for global wealth, population, or adaptation
• Modelled harm is not the same as observed harm—and the IPCC admits it
The AI adjusted. It acknowledged that nuclear is the only upward step on the energy-density ladder. It admitted fragility. It re-evaluated trust in climate models. It began, finally, to reason from observed history, not abstract simulation.
And in that process, I learned something—not just about energy or climate policy, but about arguing well in the AI age.
A Caution and a Call
Most people will not argue with their AI. They’ll ask questions and receive clean, confident answers—often reflecting elite values dressed as neutral fact. The danger here isn’t intentional deception. It’s the soft tyranny of embedded assumptions.
But there’s also hope. With the right framing—and the humility to test your own reasoning—you can turn AI into a kind of Socratic sparring partner. It won’t give you truth. But it can help you fight for it.
So I offer this as both encouragement and challenge:
AI is not your enemy—but it is not your priest either.
Ask it questions. Challenge its axioms. Demand reality over rhetoric.
Refuse to surrender your thinking to the simulations.
Use it not to confirm, but to refine.
That’s the task of the intellectually awake in an age of hallucinated certainty.
Brad - what a fascinating exercise and observation - thank you. If I may offer some reflections on AI Large Language Models (LLMs) in return.
I find LLMs both exciting and worrying. I subscribe to- and use them routinely in my own research. I understand how LLMs work and am still amazed on a daily basis by what they are capable of producing. And I also find them to be systematically infected with the biases contained in their training material, amplified by the selection and reinforcement during training of that material made by their ideologically motivated creators.
Some LLMs are also comically misconfigured. I have a keyboard macro with words to effect “Do not lie" that I append as context to every query. Without this, Claude, for example, will simply make things up in an effort to avoid displeasing you, and shrug its virtual shoulders when you confront it with having done so.
So, what do do? The test that anyone can submit them to is to examine their output on a subject with which they are already deeply familiar. To be, so to speak, an informed consumer. I mention this not to brag, but because it is indispensable to my point: I have degrees with distinctions separately in electrical and power systems engineering, petroleum engineering, and energy economics, together with 30 years of operational experience in a senior capacity in the energy sector. I am therefore a subject matter expert in a number of topics relevant to energy, climate, and economics - all disciplines in which no one person can master all topics and claim overriding authority.
As an intellectual exercise, I will therefore from time to time see if I can get an AI model to recognise the non-sequiturs inherent to the orthodox narratives about climate, "renewable" energy, etc. on which it has been trained.
This is perfectly feasible. While the model is free initially to compose its response to a query, each subsequent response takes account of earlier ones - “context”, in the jargon of LLMs - and eventually it will usually concede that the farrago of mutually contradictory statements that defending these narratives requires are irreconcilable.
Not always, though. Gemini Pro 2.5 recently cheerfully assured me that the community of 100's of thousands of people involved in studying bad weather would welcome the destruction of their livelihoods in the event that the falsification of their hypothesis was accepted, because it would symbolise the advancement of science and offer one of them the opportunity of securing a Nobel Prize. I could not dissuade it.
So if LLMs are wilfully wrong on matters with which you are familiar, then it is rational to suppose that they are also wrong on matters with which you are not familiar, and govern yourself accordingly. I only use LLMs that provide citations, treat LLM output as I would any other literature source, and evaluate each primary citation before accepting a claim on a topic with which I am not expertly familiar.
Two things worry me. Not everyone treats LLMs with this degree of suspicion or rigour. The Dunning Kruger effect has a powerfully corrosive impact on reasoned debate at the best of times. Now, the supreme confidence that often accompanies the lack of awareness of the limits of one’s own knowledge on a topic is turbocharged by a language model that appears to confirm that one’s confidence is warranted. I see debate becoming less informed, not more informed, because of this.
The other thing that worries me is what will happen when passive consumption and ignorant confidence collides with the moment when the large language model snake starts to eat its own tail - when the corrupted output from the model becomes the content on which subsequent iterations are trained, producing a kind of insanity.
The antidote? Steve Jobs once said that computers were bicycles for the mind. LLMs more so. But you still have to pedal, and you still have to steer. Purge “consensus”, “denial”, and ad hominem fallacy - pseudoscience’s tawdry hallmarks - from discourse. Demand Evidence. Logic. Reason. Open mindedness. A willingness to change one's opinion if the evidence, or your understanding of the evidence, changes (I was once Marxist-adjacent - one for another tale, perhaps).
Above all - civility. We still have to live with each other even when we disagree and, in many ways, that is what makes the intellectual life so enjoyable.
On which note - thank you again for sharing the fascinating insights arising from your analysis of this debate.
r.
This is exactly what I’ve found over and over.
However, I will often set the tone first by asking for the relevant ai to challenge my base assumptions. But from the philosophical framing of classic Christian thought.
As you note from an energy perspective I also have to consistently push for it to prioritise real world data over models. And for physical limits to always be prioritised, pulling in the thinking style of Vaclav Smil.
But normally the most useful aspect is that I tend to have very disordered thoughts and simply dumping them in one place and using ai as a tool to order love you own thinking then doesn’t outsource the thinking. But it instead sharpens what was there.
So this is a passable comment from me. But if I took time I could better structure this and pull in wider thought with ai help.
Thanks, Brad. I appreciate your insight - especially your encouragement to stay intellectually awake in the AI age. That’s sound advice in any context.
A couple of thoughts in response, since this does touch on the core of the exchange I’ve been having with Richard.
You emphasize that LLMs can reflect the worldview of “a narrow elite consensus.” But although any knowledge system has a context, I’d caution against framing that as inherently suspect. Much of what we refer to as “elite consensus” arises from the consistent findings of scientific institutions working across nations and political systems, over decades of peer-reviewed research. The danger comes not from acknowledging this body of knowledge - but from suggesting that every resistance to it is a form of independent thinking.
More importantly, I’d argue that what Richard and I were debating wasn’t LLM “bias” at all—it was the weight of evidence. I’ve used AI to help organize and clarify arguments, yes. But when I cite real-world data, observed trends, IPCC assessments, or published meta-analyses, that’s not deferring to the model - it’s actually engaging the science.
There’s a tendency now to treat any answer that aligns with mainstream climate science as “ideologically embedded” or “rhetorically polished,” while giving greater weight to contrarian views simply because they feel less institutional. But that, too, is a kind of bias - one that risks mistaking outsider status for truth.
I like your call to challenge assumptions but let’s challenge all of them - including the instinct to assume that the model is the one being ideological when it disagrees with us.
Thank you for that careful reply.
One thing that I didn’t put in the above post is that ai is built with guardrails. And these guardrails show up in interesting ways.
Often when I see a news story that I don’t know enough about I will ask ai to summarise and to steel man both sides of the issue to make it easier for me to make a clear assessment of the issue (not that I’m going to do anything with it :) ).
This works very well for most topics. But if you start to ask these questions around topics which are somewhat controversial. Or where they would run counter to the core ideological views of your average Silicon Valley progressive that’s where things get interesting. Multiple times I have done this and each time when I ask a position to be steel manned on both sides I get a very good response on the view that aligns with Silicon Valley biases. But the argument that I get for the other side is much weaker than it should be. It’s not straw man. But major foundational arguments are missed. It is only after many challenges to core assumptions that I mange to get a reasonably balanced argument.
I don’t think that the ai is suspect. It is slightly different from that. It is a mirror that will reflect and refine your own views. This means that for someone who tends to lean somewhat contrarian I run the risk of over discounting. But it may be that for you (and please forgive the assumption), you may not be sufficiently critical of your own underlying assumptions.
So looking up at the above thread I have had these same conversations with friends. Both of you are necessarily speaking past each other. And this is because your core foundations are fundamentally different.
Like you I will end up referring to ai. As I have the essence of an idea on the tip of my tongue but can’t fully articulate it yet. I’ll reply to this shortly (these are my unfiltered poorly structured thoughts).
> You’re happy to use AI when it serves you, but when I do it, you demand a comprehension test?
Correct. Large language models can be a useful research tool. But, as your exercise in cut-and-past demonstrates, they regularly produce nonsense. So to use them, you have to have an understanding of the topic which you appear not to have. In any event, passing its output off as your own is simply dishonest.
To take your model's first claim as an example. I note that there are no examples in history of a successful transition from a higher to a lower level of energy-per-person. To validate the claim that my framing is false, you would have to provide an example of a transition that was complete, that was successful, and in which energy-per-person had fallen. Your language model has selected the current transition as an example and used it to make three claims (without evidence) about renewable energy. Since our transition is not complete and, by definition, is not successful, and since increasing hydrocarbon consumption is causing energy-per-person to continue rising, the model's argument is a self-evident non-sequitur. Had you understood my point and your model's answer, you would have realised this and not pasted it into the comment box.
I'd be happy to explore its other errors with you as an exercise in understanding the extent to which large language models merely reflect the material they have been trained with. But, again, I must insist that you read the model's output first, evaluate its claims, and express the points in your own voice.
Allow me, if you will, to give you some guidance as you do this to improve your arguments.
In your claim about the role of CO2 relative to other forcing functions, it will not be sufficient for your answer to address only CO2. You will have to address all the other forcing functions, and the interactions between them, and demonstrate why CO2 is not minor relative to them over the timescale in which statements about the climate are valid.
In your claim about climate sensitivity and feedbacks, you will have to acknowledge that the only basis for forming hypotheses about them is with computer models. Then you will have to explain how these models lead to valid hypotheses about climate sensitivity and feedbacks given that the IPCC acknowledges that, as a coupled, non-linear system, the climate is computationally non-representable.
In claiming that the IPCCs findings are distorted, you will want to demonstrate that you are aware that the IPCC issues separately a political report ("Summary for Policymakers") that is written by non-scientists and, months later, a scientific report by scientists, and that these reports routinely conflict. Then you will want to be able to justify your decision to select statements from the political report about supposed human-caused climate effects and reject statements from the scientific report conceding that a human signature in these effects is not detectable.
Your model's claims about "conspiracy framing' and "ideological projection" are, I would suggest, not worth defending and, in any event, only worth attempting after you've overcome the problems in its previous claims.
I hope you'll forgive me if I run your answers through AI detection and suspend the debate if you continue to misrepresent language model outputs as your own views.
Over to you.
You’re setting up a pretty one-sided standard here. I’m expected to unpack climate models, feedbacks, and IPCC editorial structures just to question your claim that CO₂ is a minor forcing or that Net Zero is some kind of reckless downgrade. Meanwhile, you haven’t offered much to defend those claims beyond framing them as self-evident.
Let’s take your energy-per-person example. You say there’s no precedent for a successful transition to lower energy-per-person. But Net Zero isn’t a completed transition—so of course it doesn’t meet your “successful” standard yet. That doesn’t invalidate the point that clean energy can support prosperity. In fact, some developed countries have already seen energy-per-person plateau or fall while improving quality of life—Japan and parts of Europe, for instance.
You also ignore that clean energy isn’t just about raw input energy; it’s about how efficiently we use it. Electrification and efficiency gains often deliver more useful energy services with less raw energy input.
As for CO₂ being a minor forcing—how is that consistent with the weight of attribution science? The point isn’t how CO₂ compares to other natural flows in raw size—it’s that our emissions are enough to shift the balance, raise atmospheric concentrations, and trap more heat. That’s exactly what’s happening, and models, observations, and attribution studies all support that conclusion.
You brought up the IPCC. If you’re going to argue that the “scientific report” contradicts the Summary for Policymakers, I’d need to see the actual quotes and context. I’ve read enough IPCC material to know the scientists don’t back your position that human impact is undetectable. You’re cherry-picking a talking point that’s been used by skeptics for years.
And on models: yes, the climate system is nonlinear and complex—but that doesn’t mean models are useless. That’s a strawman. They’re tested, calibrated, and used across multiple lines of evidence. If you’ve got a specific mismatch between model projections and real-world data, cite it. General skepticism about complexity doesn’t cut it.
Bottom line: I’m here to engage on substance. If your arguments are solid, they should stand up to the kind of pushback I’m offering—whether I wrote every word from scratch or used a tool to help frame it clearly.
Clifton - you chose to comment on my essay, and you chose to describe it as “climate denial in a new suit: scientific sleight of hand, cherry picked studies, and conspiratorial rhetoric”.
The standard I’m setting up is the one where you defend your claim. The issues you are expected to unpack are the ones you chose in support of it. I don’t need to offer much to defend a copy-and-paste large language model dump, and I’ve coached you on how you might improve that dump to produce an critique that might need to be defended.
To your points.
Your failure to rebut my observation that there is no precedent for a completed, successful transition to lower energy-per-person is not rescued by repeating your example of a transition that is not complete. As an exercise, I suggest you obtain the data from “Our World in Data” on country size, country energy consumption, and country life expectancy. For each country, compute the ratio of energy consumption to population (energy per-person) and plot it against life expectancy. You will observe an almost straight line. Life expectancy - a straightforward proxy for “success” - falls with falling per-capita energy.
The claim that “clean energy can support prosperity” has no evidence to support it. In the case of wind and solar, the claim is physically illiterate - see my essay on this substack “The physics of Net Zero”. In the case of nuclear, it is at best aspirational - the great majority of the global industrial manufacturing system and food system on which our prosperity depends does not run on electricity and cannot as far as we know be converted to it.
Your claim that clean energy is about efficiency contains the implied claim that efficiency can be increased at least as fast as net energy is reduced. Efficiency typically increases at one or two percent per year. Net energy - the energy available to us after reserving that fraction ‘x’ of output necessary to maintain the energy system itself - falls in a 1/x relation as ‘x’ rises, creating what is referred to as the “energy cliff” or “death zone”. Solar and wind, when the manufacturing and operational requirements of their manufacturing systems are properly accounted for, are well inside the death zone implying necessary efficiency gains orders of magnitude greater than have ever been achieved. There is no support for your claim.
You claim that your hypothesis about CO2’s supposed harmful properties can be rescued through models which are “tested and calibrated”, and something derived from them called “attribution science”. There are (at least) two problems with this.
The first problem is that the IPCC itself notes that, as a coupled non-linear system, the climate cannot be modelled and, therefore, that the long-term prediction of future climate states is not possible. This is not “general skepticism of complexity”. In making this observation, the IPCC are not claiming that it is not possible with current computing power or algorithms. They are noting an insurmountable property of coupled non-linear systems, and a fundamental limitation of computational representability.
It is worse than that. For a model to have predictive capability over a physical phenomenon, it has to be a physical model. That is: not a statistical model. For it to be a predictive physical model, three things have to be true: we have to know all of the drivers of the phenomenon; we have to know the physics of each of the drivers; and the resulting system has to be computationally representable. Since none of these conditions are met, what you are referring to are not physical models.
Climate models are, in fact, over-fitted calibrated statistical correlations with tuning parameters arbitrarily determined by motivated modellers, from which no physical insight can be obtained for the observational period for which they are calibrated. From this, it follows that your “attribution science” is, in fact, a pseudo-science.
Calibrated statistical correlation models lack the ability to produce estimates outside the period for which they are calibrated. This is trivially observable. When these statistical models are initialised to 1980 and run forward (so-called “back casting”), the models with assumptions consistent with your climate catastrophe hypothesis overestimate observed temperature. Since they cannot predict today from yesterday, there is no basis for claiming that they can predict tomorrow from today.
The second problem is the question “So what?”. Suppose it was true that human emission of CO2 acted as a control knob on temperature that somehow dominated all the other climate control knobs. The people world wide who die from cold each year still vastly outnumber the people who die from heat as we emerge from the last cooling period, and orbital mechanics are already taking us into the next one. The UK Office for National Statistics computed the change in temperature related deaths in the UK over the last 20 years as a result of a very modest increase in temperature. 555,000 fewer citizens died than might have been expected had nights not got slightly less cold. So what if CO2 has a warming effect?
If you need to see actual IPCC quotes, and especially if you want to accuse others of “climate denial in a new suit: scientific sleight of hand, cherry picked studies, and conspiratorial rhetoric”, then I suggest you read them. The IPCC state that there is no observable relationship between human activity and most of the climate phenomena selected by climate catastrophists as indicators of human induced catastrophe. This is not “cherry picking”. This is “picking”.
In conclusion. I believe my arguments have stood up to the kind of pushback you are offering, and that they are solid. Thank you for the courtesy of taking the time to make it yourself.
Let me say up front: yes, I’ve used a language model here—to organize, cross-check, and sharpen my response. But these are my points, and I stand behind them. Dismissing arguments based on how they’re drafted rather than addressing their content is just a rhetorical dodge. If the argument holds, it holds—regardless of the tool used to write it.
You asked me to defend my claim that your essay reflects “climate denial in a new suit.” I’ve done so, and while you insist I haven’t, you haven’t actually addressed the substance of what I raised — only repeated your framework as if repetition itself settles the matter. Let’s take your points seriously for a moment, though.
You argue that no precedent exists for a “successful transition to lower energy-per-person,” and from this absence you imply it cannot be done. But an absence of historical precedent isn’t proof of impossibility — especially in a world that’s never faced the specific combination of existential risk, technological capacity, and global interdependence we face now. Transitions are already underway in many regions — whether measured in grid decarbonization, reduced energy intensity per unit of GDP, or improved energy services per unit of input. These aren’t hypotheticals — they’re observable trends. The fact that global energy-per-person is still rising is not a counterargument; it’s simply a snapshot of a process in motion.
You then claim that “clean energy can’t support prosperity,” calling it “physically illiterate.” But this overlooks the realities on the ground. Solar and wind are now the cheapest sources of new power generation in most of the world, and nations like Denmark, Portugal, and parts of Australia are already operating at high levels of renewable integration. Are these societies less prosperous because of it? You shift the goalposts by pointing out that many industrial systems aren’t yet electrified — which is true — but also irrelevant. Nobody claims the energy transition is complete; the claim is that it’s feasible and underway.
On the question of efficiency, you introduce the “energy cliff” argument and present it as if the entire transition collapses under thermodynamic law. That might sound impressive, but it rests on questionable assumptions. Lifecycle analyses of solar, wind, and storage systems show net-positive energy returns that continue to improve, especially when measured over their lifespans and in the context of maturing supply chains. No serious decarbonization effort is based on wind and solar alone anyway. Your portrayal of efficiency gains needing to offset all system losses at impossible rates is a strawman — nobody credible is proposing a transition on those terms.
Your model critique is another overreach. Yes, climate is a complex, coupled nonlinear system. No, that does not mean models are useless or unphysical. In fact, climate models have successfully predicted long-term warming trends, polar amplification, stratospheric cooling, and shifting precipitation patterns — all observed. Are models perfect? Of course not. But the difference between uncertainty and unreliability is meaningful. You don’t discredit a weather forecast because it can’t tell you next Tuesday’s exact temperature at 2 p.m.
You misrepresent the IPCC’s language about unpredictability in nonlinear systems as if it invalidates the entire scientific foundation of climate projection. But the models are evaluated, compared with observations, and continuously improved — not “arbitrarily tuned” to fit ideology. Attribution science isn’t a pseudoscience; it’s a statistical framework applied to observed data, peer-reviewed, and accepted across disciplines. It doesn’t need to predict the future to demonstrate that today’s extremes are being made more likely and more intense by greenhouse gas emissions.
Finally, your “so what?” argument — that CO₂ might warm the planet but that could be good — is both ethically and empirically flawed. Heat deaths are rising, food systems are being stressed, and weather extremes are growing more damaging. And you offer a temperature-mortality tradeoff as if climate were just a global thermostat. We’re not dealing with linear shifts but with tipping points and feedbacks. It’s not the average that kills — it’s the extremes, and those are intensifying.
You cite a speculative benefit — fewer cold deaths — while ignoring that the bulk of warming impacts will fall on the poorest, least responsible, and least adaptable populations. “So what?” is not an answer to a planetary risk. It’s a dismissal.
You asked for a defense of my original critique. I’ve provided one. I’ll leave it to others to judge whether your arguments rise above contrarian rhetoric dressed in technocratic clothing. But if you’re going to frame your position as reasoned dissent, don’t be surprised when others point out how much it resembles the very denial you think you’ve transcended.
So you really don't understand what the problem is.
Keep working on the insults, though. Practice makes perfect.