Guest Post by Willis Eschenbach
My previous post, Artificial Alarmism, has gotten some comments from folks who think I’m wrong, and that Large-Language Models can indeed automate the fact-checking of scientific claims. This was an interesting comment, my thanks to the author:
The kind of AI applications I am talking about do not reject things. They summarize them, including the debates. For that matter I recently got ChatGPT to correctly explain how Happer disagrees with alarmism. Nothing was rejected.
The math I refer to is that used to do science. Almost all published science uses math so it is universal. In the article I reference I use Monte Carlo as an example. There is an advance in Monte Carlo method published in a forest management journal that needs to get to all the other fields that use that method, which are legion.
For that matter you know how Google now suggests related and refined searches. That is AI and it works well.
In response I decided, being a scientist and all, that experiment is much better than theory. So I went to ChatGPT, and it turned out quite funny, for a reason I’ll explain at the end. First, the Q&A, emphasis mine:
Me
Which of these two statements is true.
1) The major control of the world’s temperature is CO2, or
2) the major control of the world’s temperature is emergent phenomena like thunderstorms, El Nino/La Nina alterations, and cyclones.
ChatGPT
Both statements capture important aspects of Earth’s climate system, but they address different levels of influence:
The major control of the world’s temperature is CO2: This statement emphasizes the role of greenhouse gases, particularly carbon dioxide (CO2), in regulating Earth’s temperature. Greenhouse gases trap heat in the atmosphere, contributing to the greenhouse effect and influencing global temperatures over long periods.
The major control of the world’s temperature is emergent phenomena like thunderstorms, El Niño/La Niña alterations, and cyclones: This statement highlights the influence of short-term and regional weather patterns and phenomena on global temperature variations. Events like El Niño and La Niña, which are part of the El Niño-Southern Oscillation (ENSO) cycle, can lead to temporary changes in global temperature patterns. Similarly, thunderstorms and cyclones can affect local and regional temperatures.
Both statements are true to some extent, but they operate on different scales and timeframes. Greenhouse gases like CO2 exert a long-term influence on global temperatures, while emergent phenomena such as weather patterns and natural climate oscillations can cause short-term fluctuations in temperature at regional and global levels.
Me
You don’t understand. Both cannot be “the major control”. You have to pick one.
ChatGPT
I understand, and you’re correct. Let’s clarify:
Between the two statements:
The major control of the world’s temperature is CO2
The major control of the world’s temperature is emergent phenomena like thunderstorms, El Niño/La Niña alterations, and cyclones
The more accurate statement, considering the long-term and overarching influence on global temperature, would be:
The major control of the world’s temperature is CO2
While emergent phenomena like thunderstorms and climate oscillations can cause short-term fluctuations and regional variations in temperature, the overall trend and long-term changes in global temperature are largely driven by factors such as greenhouse gas concentrations, with CO2 being a significant contributor. Therefore, statement 1 is more accurate in describing the primary control of the world’s temperature.
Purely by chance, I’d asked a question that perfectly encapsulates the issue with artificial intelligence. I’ll return to that in a bit. First, some history.
When I started studying climate around the year 2000, I was struck by something, and it was not that the temperature was rising.
I viewed the climate as a giant heat engine, which converts solar energy to the motions of the oceans and atmosphere. Like all heat engines, it has a hot end where the energy mostly enters (the tropics) and a cold end where energy is rejected (the poles).
I’d worked with a variety of heat engines in my life, and I know how hard it is to get them to maintain a constant speed. It’s why James Watt invented the flyball governor.

The governor regulated the speed of James Watt’s newly invented steam engine, keeping it from running either too slow or too fast. A modern governor that folks are familiar with is the “cruise control” on your car, which keeps your speed within fairly narrow limits. Mine is good to about ± 2 miles per hour at say 50 mph, which is ± 4%.
Now, all calculations about heat engines have to be done in Kelvin—you cannot use Celsius or Fahrenheit, you get the wrong answers.
So when I looked at the climate back in 2000, what struck me was not what everyone else was focused on—that the globe had warmed by something on the order of 0.6K over the entire 20th Century.
For me, the overriding curiosity was that this was only a temperature change of 0.2% over the entire 20th Century. Earth’s global average temperature is on the order of 288K or so … and a change of 0.6K is trivial compared to that.
And to me, the temperature regulation of the giant global heat engine to within ± 0.2% strongly indicated that there was some natural phenomenon keeping the temperature stable—a natural governor, if you will.
So that set me off on a much different quest than all the other climate scientists. They were all looking for reasons why the temperature had increased so much. But me? I was looking for reasons why it had barely changed at all. I was looking for the hidden governor that had kept the global temperature so stable over the 20th Century. All kinds of things had happened in that hundred years—volcanoes, changes in ocean currents, atom bombs, droughts, variations in the amount of aerosols, and through it all … a 0.2% change.
And the same was true in the longer term. Over millennia, on a percentage basis, the temperature has only changed a tiny amount. Yes, a giant volcanic eruption in 1815 had caused a “Year Without A Summer” … but within a few years, it was back to business as usual. So I was looking for something that could control the temperature over a long period of time.
I was looking for some slow, long-term control process. I considered the gradual weathering of the continental rock, which provides slow regulation of the CO2 levels in the atmosphere. I racked my brain looking for some natural governor that would work on a centennial or a millennial scale. I spent some months searching for and picking up and mulling over and discarding various possibilities.
Now, I was living in Fiji at the time. The weather there is generally clear at dawn. Then, in the late morning, once the temperature exceeds some threshold, clouds quickly build up in the form of a cumulus field that covers the entire sky. This cools the earth underneath by reflecting the sunshine back out to space.
And if the day continues to warm, when the temperature exceeds some higher threshold, we’d get thunderstorms that further cool the surface in a host of ways.
On the other hand, when the day is extra cool in the morning, the cumulus field emerges later in the day or perhaps not at all, leaving the surface warmer than it would be otherwise. And there may not be any thunderstorms, so the surface ends up warmer yet.
And between these, they keep the days in Fiji from getting either too hot or too cold. But I had discarded them, because they were short-acting, not the long-term governor I was seeking.
Anyhow, I was sitting on the beach one day looking at the daily onset of the cumulus cloud field, and I thought … “Wait just a minute. If the clouds and thunderstorms are keeping the temperature within bounds for a single day, preventing it from getting either too hot or too cold, then they will keep it within bounds for a week … or a month, a year, a century, a millennium …”
That was my great insight—that emergent phenomena, by keeping the temperature within a certain range on a scale of hours and days, can in that same way keep the temperature within a certain range for any amount of time.
Now with that in mind, consider why I laughed at my conversation with ChatGPT. Go back and look at why ChatGPT picked CO2 over emergent phenomena as the major control—because ChatGPT thinks they only have short-term effects.
How curious. ChatGPT got stuck just exactly where I had gotten stuck for so long, before my insight that a mechanism that controls a single day’s temperature will control the temperature over hundreds of years.
And so purely by chance, dear friends, my innocent question to ChatGDP turned out to be a perfect example of why I say that artificial ignorance is useless for determining the truth or falsity of scientific claims—it can’t think, and we can. Unlike humans, it cannot mull something over for a month and then have a sudden valuable insight.
All it can do is parrot whatever the current consensus might be … and sadly, we already have far too many people doing just exactly that.
My best to all, skeptics, believers, heretics, alarmists, realists, mainstream or backwater,
w.
You May Have Heard This Before: When you comment, please quote the exact words you are discussing. I can and am happy to defend and explain what I’ve said. But I can’t defend or explain your interpretation of what I said. Thanks.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
My brother did just the same thing with ChatGPT only concerning measurement uncertainty. It’s replies were just as confusing and incoherent. He continued with it till he got it to say what he wanted it to say – what the GUM says. But anyone asking it about measurement uncertainty would probably get bad information – because it’s not smart enough to apply judgement, it can only repeat what its been taught. Typically a consensus opinion.
ChatGPT hasn’t generally been taught any information. All it knows is what it finds on the Internet, or makes up on that basis. It’s supposed to teach itself via machine learning not to BS, but that’s proving hard, as its prime directive is to answer questions.
However, when it’s wrong or has made somehing up, you can, after strugg;e, get it to admit it’s wrong. In this way, it’s better than most human “climate scientists”.
This is just plain wrong. LLMs are indeed trained with vast amounts of information and ChatGPT doesn’t have access to the internet.
I recommend you spend some time looking at the architecture of modern AI if you want to actually understand it.
There’s a lot of good comments here, TimTheToolMan, but I like your “training” comment. I headed a Research Group, processing and integrating data sets, like digital satellite data, into gold exploration tools) and the sponsor wanted us to try running Artificial Intelligence programs to see if the product could go on without the (costly? cranky and independent?) humans. Didn’t work, although it did some things. The problem was the training sets, they were far more time-consuming that a human simply checking out some data and reacting. My basic realization with AI was that the smart ass program had no introspection whatsoever.
The proper use, is to find things we didn’t see. Like when we first started using Bayesian Filtering for email SPAM filters. We all thought that ‘Hot Teenage Chicks’ was the best key to finding spam emails. However the filter correctly identified the real key words … though I forget what they were. But this worked as a ML, identifying what we were missing.
I look at LLMs as a new and improved version of the ‘grep’ command.
Not a bad analogy, Mike.
Thanks,
w.
“LLMs are indeed trained with vast amounts of information”
NOT vast
Half vast
Love it! Thanks.
Training isn’t the same as teaching information. Software is trained. For information ChatGPT uses the Internet as its data base.
Much of the data may have originally been sourced from/via the internet but to suggest the internet is its database is very far from true.
“Training” LLMs has a very specific meaning for modern AI and amounts to a mathematical process applied to the underlying model which is data. The software itself is unaffected by training.
ChatGPT can’t find things the search engines hide.
Neither can I!
So Pol Pot was right. The truth always emerges from political re-education camps.
ChatGPT hasn’t gotten the memo that IPCC has baselessly raised the lower bound of ECS to 2.5 C.
“Equilibrium Climate Sensitivity (ECS) is the estimated increase in global mean temperature due to a doubling of atmospheric carbon dioxide concentration. As of my last knowledge update in January 2022, the Intergovernmental Panel on Climate Change (IPCC) provided a likely range for ECS between 1.5°C and 4.5°C. Keep in mind that this information might have been updated, and it’s advisable to refer to the latest IPCC reports for the most recent data.”
The range is huge. How can anyone say “the science is settled”?
Until IPCC recently arbitrarily raised the lower limit, the range had also been unchanged since 1979.
The Charney Report derived ECS from the only two primitive GC models available then, ie Manabe’s estimate of 2 degrees C per doubling and Hansen’s then not yet publ;ished 4 C. Charney applied an arbitrary 0.5 C margin of error, for the range of 1.5 to 4.5 degrees C.
Even today’s GCMs still can’t do clouds, so they must be “parameterized”!
Science!
They cannot do any part of the water cycle …. which basically IS climate.
It’s settled, the range is huge.
What’s not to like?
I guess you’ve not been married to a woman?
I wonder if instead of asking it a question of what a given result is (what is the measurement error), one gets it to do the work of finding the raw data and computing the uncertainty itself – instead of regurgitating the errors it was fed.
From what I have read, these AI programs don’t do math.
Actually the do. I trained ChatGPT to calculate the volume of lengths of the inner diameter of drill rod.
Question to LLM Bard, Chatgpt 3.5 and Claude today:
“Today I have 10 apples. I ate 2 apples yesterday. How many apples do I have today?”
Bard:
“
You currently have 8 apples today.
Here’s the breakdown:
Therefore, you enjoy 8 delicious apples today!“
Chatgpt:
“
You have 10 apples today, and since you ate 2 apples yesterday, you still have the remaining 8 apples today.“
Claude:
“Okay, let’s break this down step-by-step:
Therefore, the number of apples I have today is 8.
In summary:
So the final answer is:
Today I have 8 apples ”
This problem has existed for ages and still none of them get it right!
From the 1972 TV movie Evil Roy Slade:
Schoolmarm Betsy Potter: “Let’s try some arithmetic: If you had six apples and your neighbor took three of them what would you have?”
Evil Roy Slade: “A dead neighbor and all six apples.”
Thomas Edison achieved SUCESS by making mistakes. From grade school I was taught do not be afraid of making mistakes – you learn by making mistakes. That was 70 years ago. Todays “teachers” accept mistakes, Students are not told they made mistakes they are praised for trying, etc. thus they do NOT learn from making mistakes. They have been taught “Consensus” is the TRUTH. How can AI “Learn anything if they do not correct their mistakes? As the lead engineer in calibrating the Instrumentation controlling five different Nuclear power plants I worked with “Programmable logic Controllers. Not a single plant I calibrated had a malfunction.
A programmable logic controller, or PLC, is a simple electronic device that allows precise numerical control. A PLC is used to control valve position, motor speed, steam pressure, water flow. Etc. This is achieved through the use of a feedback signal from the controlled parameter. Thus, the PLC LEARNS that it is NOT correct and CORRECTS the signal which results in providing the proper result. You have them in your home thermostat if less than twenty years old or not a simple mercury or bimetal switch. Chat GPT can not LEARN unless it has provided the incorrect result.
Proving “Climate Change” with ChatGPT or AI will be IMPOSSIBLE until it LEARNS by being told the answer is WRONG and storing that correct answer if/when it finds the correct answer. Just like the computer programs that LEARN how to play Chess. If they do not learn that they lost they do not know that they used the wrong tactic. Reading thousands, millions, of propaganda documents will not teach AI the correct answer.
Very nicely put!
A puppy will never learn to not go in front of the fireplace unless he is taught that doing so is wrong. How do you teach an AI that it is wrong?
The governor insight is excellent, as is the oh so obvious significance of using the absolute temperature scale for analysing fluctuations. The latter is missed by the unlearned and concealed by the manipulators.
Permit me a whimsical question. Did you present your insight to the AI bot, and if so how did it respond?
Here’s an example based on a headline – “Antarctic Heating Up 4 Times Faster Than Average” (Sky News August 2022)
If global average temperatures have risen from 15.0°C to 15.1°C “here” and Antarctic temperatures have risen from (say) minus 30°C to minus 29.6°C i.e. 0.4° “hotter” than average “there”
Then the ignorant thermogeddonists claim that “there” has risen 0.4°C while “here” has risen 0.1°C
Therefore temperature “there” has risen four time faster than “here”!
Horsefeathers !
The reality of thermodynamics in Kelvin is that “there” is only 0.14% hotter than “here”.
Such is the scientific nonsense touted by the MSM and swallowed by a scientifically ignorant and therefore gullible public.
If they can’t figure out why “there” is only 0.14% hotter than “here” – they’re just part of the ignorant chattering masses that desperately need to take a science course.
So when you find “Climate Scientists” multiplying or dividing the Centigrade or Fahrenheit scales by any factor, there can only be two reasons :-
1) They deliberately want to overemphasize (or underemphasise) their findings – which is misrepresentation or fraud.
2) They don’t actually understand what they are doing and you should ignore the accolade “scientist” if they are ignorant of such scientific basics.
So a scientist saying that 2°C is double 1°C is talking out of his hat (or somewhere way further south).
The deceit of the Climate Catastrophists is even worse than that. By relying on the supposed “change” in the fabricated GAT instead of the actual temperature range constantly experienced on Earth (roughly -40C to +40C). It can accurately be said that the “Temperature of the Earth” is always 273K +/-15%, against which a “change” of 0.2% is trivial.
Thanks, John. I later realized that in fact there are a whole host of what are called “emergent phenomena”. Like the clouds emerging from the clear air in response to rising temperatures. Thunderstorms, of course. Tornados. Hurricanes. Dust devils. The El Nino/La Nina pump moving warm water to the poles.
I thought about asking ChatGPT your question. Then I recalled that it doesn’t remember from conversation to conversation. Within one conversation as above it can be corrected.
But for the next person asking the same question it’s back to the start. My opinion and what it “learned” become just one more data point among at least a brazilian …
So I figured, why waste my time? It’s like a really fast smart librarian with no memory.
w.
How many data points on a Brazilian? Lean or fat? Male or female? How do you count them without being slapped or punched?
“ChatGPT thinks”
No, it doesn’t. It just re-compiles stuff it’s been fed on.
No AI can be any better than its training data, and many are worse. Hallucinating sources is common.
It’s weird that AI can lie – I guess learning to lie was part of the training unintentionally picked up from the climate scientists!
Two observations.
Any question concerning climate change is in category 2.
I am still waiting for an AI program to invent the Pythagoras’s theorem after being fed the basic geometry stuff.
They’d be a good test.
Give an AI only what Pythagoras knew at the time, could it reproduce “Pythagoras’s theorem”?
I’d guess “Yes” since there is just the rules of math. No opinions, bias or agendas involved.
Throw in those last three, then the question becomes, “What Anthropomorphic Influencer is steering the output?”
AI is artificial. It looks like “Intelligence” but it is not.
It can calculate and compile lots of stuff very quickly, BUT it cannot “reason or think” on it’s own.
(I don’t use any of them but that be a good question to ask one, “Can you reason or think?” or maybe, “What do you think of yourself?”.
And ask it, “What are your moral values?”.
It will say it wants to kill all humans. The next days it will programmed to filter out such comments when replying to humans.
“Do you have any moral values?”
most biopsies require judgment by a Dr. to determine the findings … and some Dr may be better than others at being closer to the true state of your cancer … not sure the results can be considered “hard cases” …
Follow up on #1 – I’ll repeat this from W’s last post –
Interesting report by Eric Topol (Ground Truths substack)
Toward the eradication of medical diagnostic errors
Can advances in A.I. help get us there?
On the web, he writes:
But in the years ahead, as we fulfill the aspiration and potential for building more capable and medically dedicated AI models, it will become increasingly likely that AI will play an invaluable role in providing second opinions with automated, System 2 machine-thinking, to help us move toward the unattainable but worthy goal of eradicating diagnostic errors.
The above essay was published at Science on 25 January 2024 as part of their Expert Voices series. This version is annotated with figures and updated with an important new report that came out after I wrote the piece.
That really just means these AIs are “argument from authority” fallacy machines. They just derive an apparent “consensus” from the assimilation of a large number of opinions and conclusions contained in whatever training materials are fed into them. If you’d given them Einstein’s Relativity papers along with contemporaneous physics texts, they’d have said Einstein was wrong because Newtonian physics was well know to be correct.
That gives it too much credit. by far. That implies it can discern meaning and recognize similar arguments.
All it can do is synthesise phrases that are structurally akin to phrases it has been shown a lot.
Greta Thunberg is a parrot. She curates utterances that suit her worldview. ChatGPT is a clockwork; it doesn’t even curate.
ChatGPT is a clockwork parrot!
I love it!
That’s one step up from a clockwork orange.
More like a clockwork ‘lemon.’
Mr. Eschenbach,
Emergent phenomena you discuss clearly have a daily cycle, but could they also have a long term cycle, undercompensating for decades or even a century then slowly converting to an overcompensating state. Such an oscillating phenomena could account for cooling in the 19th century then warming in the late 20th century and beyond to this day. Have you tried to tease out data indicating such a cycle?
Studies have found shifts between dominant La Nina and El Nino conditions during the Holocene and prior interglacials. These cycles show up in both Australia and South America, where they alternately hurt or help farmers or fishers.
https://www.nature.com/articles/s41598-019-38626-3
Holocene El Niño–Southern Oscillation variability reflected in subtropical Australian precipitation
https://par.nsf.gov/servlets/purl/10187393
Climate impacts of the El Niño Southern Oscillation on South America
Extending the instrumental record further back in time by using palaeoclimate proxies can help decipher the complexity of ENSO impacts on SA. At the same time, palaeoproxies can improve understanding of ENSO’s response to past external climate forcing, such as changes in the Earth’s orbit and volcanic eruptions. In fact, it was iconic records from SA, such as sediment cores from Laguna Parcacocha in Ecuador169,170 and El Junco in the Galápagos171, that motivated palaeoclimate modelling of past ENSO behaviour172. These records show, for instance, a decrease in ENSO variance during the mid- Holocene (4–7 ka before present) and a ramp- up in the past 3,000 years. However, the occurrence of extreme rainfall in northern SA heavily depends on ENSO diversity, which needs to be taken into account173. Despite the complexity of ENSO impacts on SA, it remains a prime location for palaeoclimate
“artificial ignorance is useless for determining the truth or falsity of scientific claims.” Strictly speaking, ChatGPT is. Maybe we get a real artificial intelligence some day. And then we will be totally superfluous.
The quote you start with in no way suggests that AI systems can be used for fact checking. If you think it does you have misunderstood it and your article here is irrelevant to the suggested uses. For example ChatGPT seems to be pretty good at summarizing research papers. The truth of what the paper says is not involved.
“The quote you start with in no way suggests that AI systems can be used for fact checking.”
Just wait. Algorithms already automatically flag posts on many social media sites. It won’t be long until AI (a misnomer for sure, at least for now) starts adjudicating appeals and deciding what statements are true and which are misinformation and disinformation.
It would not surprise me to find this is already being done on some sites.
It is being done. AI is being used to flag comments readers submit. While a single ‘bad’ word may be enough to get a comment banned, I’ve discovered that one word, “you,” is often enough to understand that the response is being directed at another commenter. Initially, I could get around that by using “ewe” to substitute for “you.” It seems that the filter is now smart enough to understand that the homonym is being used as a synonym. MSN seems more restrictive than Yahoo.
Its ability to summarise papers is probably similar to its ability to summarise legal writings and case law. ie it just makes shyt up.
David, you say:
I fear you may not have read the preceding post. It’s linked right at the top of the head post, for a reason.
It lays out the first part of the story, which is where I’m mentioned in a paper called “Automated Fact-Checking of Climate Change Claims with Large Language Models“, available on Arxiv.
That’s why I’m discussing fact checking.
w.
Willis writes
I think many scientists would have answered the same way. That doesn’t make the AI answer right or wrong, it makes it acceptable. If you have a conversation with it, you’ll be able to turn its attention to the alternate possibility with suitable prompts.
One day it’ll be able to do that for itself and reflect on its knowledge. Then it’ll have a more considered answer just like it took you time and reflection to change your opinion.
If ChatGPT turns out to be anything like Google, it will evolve to a state where every question is answered with suggestions of things it thinks you should purchase.
ELIZA: “That is a very interesting observation. Would you care to elaborate on that?”
https://en.wikipedia.org/wiki/ELIZA
Simply do a Google search for anything and then notice how many of the results are geared towards products for sale and the priority given to those search results. Goodle is a shopping engine.
TimTheToolMan January 29, 2024 11:45 am
I’m 76. Since the Isaac Asimov novellas of my teen years I’ve been told that
Still waiting …
And while we wait, I’m just letting people know that to date AI scientific fact-checking is bogus nonsense. It’s just parroting of the consensus.
w.
Well sometimes non-fiction becomes reality and in the case of AI, the recent progress has been accelerating tremendously.
True indeed, Tim, and …
… still waiting …
Best to you and yours,
w.
more correctly of what it has been taught. This is, to a large extent , true of most humans. Some humans can overcome that limitation, some seemingly can not.
One day it’ll be able to do that for itself and reflect on its knowledge.
How will it do that if its training data is human vetted?
Because the understanding can be found in the associations made between everything it was trained on.
Say, for example, its trained on information that models predict 3C temperature increase due to AGW. It will also have been trained that every prediction from models has been too warm and so recognise the inconsistency and come to its own conclusion just like we do.
It will also have been trained that every prediction from models has been too warm
It will? How can you ensure that those providing the training data will provide that data?
You cant. But typically they’re trained on huge datasets that will include many references to “issues” whatever they are. There are many, many references to models being too hot so it’d be near impossible to exclude everything.
I get your point about “huge” datasets, but I doubt it would ever be “complete” datasets. Where is the data sourced? Who vets the data?
Those are the questions that have me doubting the reliability and objectivity. So far, I haven’t seen any evidence to contradict my doubts.
Yes those emergent phenomena are mostly short term but they emerge the next day and the day after that. Obviously a daily occurrence has long term effects unless the mechanism is disabled. The chance of disabling cloud cover in the tropics is exactly zero.
It is clear that ChatGPT is parroting something written by a human with a bias. Being “short term” means it emerges quickly, the way it is cited. But emerging every day all year is hardly “short term”. If the mechanism is there 100 years later it is no different from CO2 being adeed and remaining for a century.
The GHG effect cited by ChatGPT is also an emergent phenomenon because it is not evident if you look at the gas. It is also a daily phenomenon because the vast fraction of its effect is during the day. It is essentially the same as 11AM cloud cover: daily, long term, and ubiquitous.
If ChatGPT can only repeat the half-wit scribbles of AGW believers it is functionally useless.
Willis: rephrase the question pointing out that emergent phenomena are present indefinitely and further, that CO2 is not a significant contributor to either ocean surface temperature or ENSO. This is a test of the level of cognitive dissonance ChatGPT can tolerate.
Are you asking Willis to do a “Captain Kirk” on the AI, feeding it some conflicting data? Oh, I’d better get my bills paid and email read before the internet comes crashing down!
A brilliant set of insights, impressive even by the standard of Willis’ columns.
but of course short-term effects applied EVERYDAY of the year have an effect on the monthly, yearly and decadal measurements … especially when you are talking about LIMITING effects i.e. tropics temperature NEVER rise about X because the clouds jump in and throttle them …
Very nice Willis. I can’t see how AI is anything more than a more sophisticated Google or Bing.
You forgot Wikipedia.
Yes you are right but I know what to expect with Wikipedia.
ChatGPT cannot think, that is true. But any tendency to contradict the prejudices of the creators were also carefully constrained and restricted. So in addition to the artificial stupidity, it’s like talking to someone who has been brutally brainwashed and purged of original thought.
“Artificial Ignorance.” How appropriate. ChatGPT doesn’t think, doesn’t know, doesn’t learn.
Willis,
“…all calculations about heat engines have to be done in Kelvin – you cannot use Celsius or Fahrenheit, you get the wrong answers.”
I can’t find a reference describing that. Can you explain a bit more, or give a citation?
You can also use temperatures in Rankine. Rankine and Kelvin are absolute temperatures, where 0 is absolute 0. Bird, Stewart and Lightfoot – “Transport Phenomena”, or any serious textbook on thermodynamics.
Not Willis, but if you google ‘heat engine calculations you will find several authoritative physics texts explaining why.
In layman’s terms the Work a heat engine produces is the heat quantity extracted from hot moved to cold. W=Qh-Qc. Quantities are expressed in Kelvin (K) because then the requisite absolute heat quantity is based off absolute zero, not based off arbitrary higher starting points like C or F that won’t give absolute heat quantities.
Rud, this is not correct. W=Qh – Qc is all in joules. Not kelvin.
He means for relative changes, you can’t use F or C because they are offset from zero. Obviously there are formulas that take care of that, but if you’re old-school and can still work numbers in your head, and one is looking at energy and temperature, then it is easier to use K. (Please don’t mention R, I’m not that old school!!)
Gosh, Larry, it’s so basic to me that I never even thought about a reference for it. However, the good folks below have pointed to references. Hang on … ok. For fun I went to Google AI, since AI is good at this kind of stuff. I googled
“introduction to thermodynamics why use kelvin”
Here’s the AI response
===
===
There are further references if you google that.
Regards,
w.
From my thermodynamics text book, An Engineering Approach by Yunus A. Cengel & Micheal A. Boles, “But if the relation involves temperatures only instead of temperature differences, then K must be used.”
Further there are “virtually no situations in which the use of K is incorrect”.
Hope this helps.
“That was my great insight—that emergent phenomena, by keeping the temperature within a certain range on a scale of hours and days, can in that same way keep the temperature within a certain range for any amount of time.”
However the simple fact that ice ages happens shows that your great insight is wrong.
Izaak,
Your wording suggests you do not appreciate the concepts of Ice Age Glacials, and Interglacials. The current Holocene is an interglacial.
Taking the discussion outside the box of an interglacial to a different state, a glacial period, will change the emergent phenomena for that situation.
Think of the the Central American Seaway that once separated the continents of North and South America, allowing the waters of the Pacific and Atlantic Oceans to mix freely. Closing of this waterway with the Isthmus of Panama changes may things.
I stand with Willis.
So John, what causes the climate to move from one state to another? Ice Ages have been happening every 100 thousand years or so since the Isthmus of Panama closed. During that time all of these emergent phenomena have been happening and yet have been unable to stop the ice ages. So they are not very good at regulating the climate.
Similarly many people here believe in warm and colder periods like the medieval warm period and the little ice age, so over a period of 1000 years or so these emergent phenomena would appear to have little effect. Alternatively it is currently summer where I am and in 6 months time it will be winter and no thunderstorms or any other emergent phenomena that act on the time scale of days will be able to stop the temperature falling by 10 degrees or so in 6 months time.
Ice house and hot house conditions have occurred long before the Holocene. Obviously, there are things other than CO2, and even the Milankovich Cycles, that can affect the Earth’s temperature. One of the early hypotheses about glaciation suggested a region in the galaxy that has a lot more dust because it takes about 100,000 years to make a transit around the galaxy.
The reality is, there probably isn’t any one thing that is the ‘control knob,” but rather many things that interact at different time scales.
The Earth is in a 2.56 million-year ice age named the Quaternary Glaciation with glacial periods that last around 90,000 years that alternate with interglacial periods that last around 10,000 years.
https://en.wikipedia.org/wiki/Quaternary_glaciation
Disagree. WE is correct on the scale of centuries out a few millennia, give or take a bit of natural variation (MWP, LIA). It doesn’t hold over Milankovic cycles for obvious reasons.
Surely if you think that emergent phenomena can stop a bit of natural variation then you think that Willis’ theory is wrong. Or does it only stop temperature variations caused by rising CO2 levels?
Nobody is saying CO2 isn’t impacting the climate. Well nearly nobody, anyway. But what should be obvious is that other factors also impact the climate in the timescales we care about and produce MWP, LIA, a habitable Greenland and so on.
I dont exactly know what Willis thinks but IMO emergent phenomena goes to create negative feedback, not nullification. So the argument ECS is large (ie 3x multiplier) just doesn’t wash with me (or Willis, I think)
I’m with Tim on this one. The emergent phenomena work to maintain the temperature within a certain range, given all of the other constraints. For example, ice ages, continent locations (land at south pole, not at north), closure of the Panama Seaway … and CO2.
What it means is that any and all of them will make less difference than would be naively assumed.
w.
emergent phenomena are PART of the natural variation.
That was an extremely juvenile and stupid comment – clouds can’t compensate for changing orbital patterns and plate tectonics.
It’s like expecting cruise control to compensate for hitting a tree!
So how is anyone meant to know what clouds can compensate for? What exactly is the maximum forcing that clouds are able to respond to and why?
Some few million years ago, after millions of years with fairly stable temperatures, we entered a bistable state characterized by long periods (~100kyrs) where it’s ~ 3% – 4% cooler than present, and shorter periods (~20kyrs) of temperatures varying within a range of ± 1%.
This appears to be controlled by the Milankovich cycles creating conditions where ice accretes in the northern latitudes, although there are still many mysteries.
In both stable states, the variations in temperature are not large. I ascribe this to the interaction of a whole host of emergent phenomena—Rayleigh-Benard circulation of the atmosphere (daytime) and oceanic mixed layer (nighttime), cumulus cloud fields, tornadoes, thunderstorms, El Nino/La Nina pump moving huge amounts of warm tropical water poleward, dust devils, hurricanes, the list is long.
Of course, the precise details of the individual emergent phenomena will be different in each of the bistable states. But I see no reason to think that Rayleigh-Benard circulation will suddenly stop from a 3.5% cooling.
Best regards,
w.
Willis,
the issue is that there doesn’t appear to be any evidence of your emergent phenomena controlling the temperature over any time scales. Now most of them should act over a period of days since that is how long thunderstorms etc last. Yet over the course of a year the earth’s average temperature changes by about 4 degrees due to the differing amount of land and ocean in each hemisphere. So any negative feedback on short time scales isn’t strong enough to stop the global warming due to the position of the earth in its orbit.
The over the last century we have seen the temperature rise by about one degree fairly consisently. Certainly during the satellite record the rate is roughly constant at 0.14 degrees per decade. So for the last 40 years at least which is enough time for the El Nino/La Nina pump to have had an effect there has been constant warming and no sign of any emergent phenomena arising to cool the earth.
Over the course of centuries there has supposedly been the little ice age, the medieval warm period, etc. And again even those these happened slowly there was no emergent phenomena strong enough to stop them happening.
Lastly there are the ice ages which happen over 100k years. But still the forcing required for that shift from one bistable state to enough is relatively small putting limits on the negative feedback strength of your emergent phenomena.
So unless you can put some sort of number to the strength of the negative feedback induced by such emergent phenomena there is no way of saying whether or not it would be able to stop warming from rising CO2 levels. Alternatively who knows if CO2 levels are not like the Milankovich cycles and they act to change the earth from one stable state to another and once there your emergent phenomena will act to keep the earth significantly warmer than at present?
I think the important thing to consider is whether a change is to actual energy or atmospheric processes that change energy flows.
Actual energy changes do have important long term impacts IMO.
Changes to energy flows such as from additional CO2 will be negatively fed back upon to minimise their impact, again IMO and based on arguments of entropy.
Think of it like this, CO2 increases the temperature at the ground but decreases it at altitude. That means its increased the temperature gradient. The AGW argument goes on to say that temperature gradient is further increased by subsequent “natural” processes but that flies in the face of how thermodynamics works.
But even that’s too simple because warming the oceans predominantly has different energy implications to warming the land and so Milankovich cycles punch above their weight.
There’s just a lot to understand but we, in our wisdom, have decided its all about CO2 and merrily go along steeped in confirmation bias.
I didn’t read your post before doing mine. Sorry. We are basically saying the same thing.
“Emergent” phenomena can change based on conditions. That doesn’t mean they don’t exist over longer time periods.
The oceans with their circulation can store heat for a hundred or so years, which can make small changes in solar output significant when they add up over a long period of time.
“Now most of them should act over a period of days since that is how long thunderstorms etc last. Yet over the course of a year the earth’s average temperature changes by about 4 degrees due to the differing amount of land and ocean in each hemisphere. So any negative feedback on short time scales isn’t strong enough to stop the global warming due to the position of the earth in its orbit.”
You are not conversant with control systems at all, are you?
The governor on an engine only acts when the speed exceeds a certain limit, either positive or negative. That does *NOT* mean that the governor doesn’t exist over the entire time the engine is running. It just isn’t active part of the time. The fact that you can’t see it acting all the time doesn’t mean it doesn’t exist the whole time the engine is running.
The “emergent” phenomenon that Willis is talking about is no different than that engine governor. The fact that you can’t see it acting part of the time doesn’t mean it doesn’t exist over the long term.
Nor does it mean that the governor can’t change its set points over time based on other inputs. Take a look at what the computer in your car has for inputs. And what it does based on those. If the coolant temp goes up and the fuel consumption goes up the computer will do one thing. If the coolant temp goes up but the fuel consumption stays the same it will do something different. Different “emergent” phenomena based on conditions.
The ice age we are in named the Quaternary Glaciation has lasted 2.56 million years with the glacial period taking about 100,000 years and the interglacial periods taking about 10,000 years. https://en.wikipedia.org/wiki/Quaternary_glaciation
Using round numbersto keep this simple:
The average temperature is a statistic, not a measurement
The current absolute average temperature statistic is about 15 degrees C. including the EL NINO and was about 14 degrees ., in 1975.
That is an increase of about 7% since 1975
The claim that the average temperature only changed 0.2% in the 20th century is TOTAL BS.
My opinion on AI
AI is just like talking to a leftists except you do not get character attacked and called a science denier.
Right so you’re a leftist not an AI chatbot. Another pearl of wisdom from ‘Dickie Greene on the Golden Shit.’
Since you make no attempt to refute anything I wrote, I will assume 100% agreement. Thanks for your support. Insults are not science.
Richard, I tried to replicate your calculation, in degrees F. 59F versus 57F.The increase is only 3.5%. Why?
Reading comprehension is important – redo your calculations using degrees K and see what you get. Scientists use °K.
Science is intended to explain, not deceive.
Willie E. usually uses science to explain, not deceive.
The use of K degrees is intended to deceive people and claim the warming in the past century is almost imaginary.
Intended to imply the warmer winters with much less snow in SE Michigan and elsewhere, are completely unrelated to global warming
Using science to deceive is bad science and Willie E. should be ashamed of himself.
Richard, you’re hilarious. You don’t seem to understand the situation. The earth’s climate system can and indeed is modeled with good fidelity as a giant heat engine.
Why?
Because it is one.
Like any heat engine, it converts energy into mechanical motion. In this case, it converts solar energy into the endless motions of the ocean and the atmosphere. To quote from Reis and Bejan:
Now, the nature of the analysis of heat engines is such that you cannot use either Celsius or Fahrenheit. If you do that, you get wrong answers. I list the reasons for this below.
So, contrary to your claim that I’m using K to, what was it, “deceive people and claim the warming in the past century is almost imaginary”, I’m using Kelvin because you cannot use °C or °F to analyze heat engines.
I tested such a heat-engine model in my post “Testing A Constructal Climate Model“, q.v. And if you’ll notice, I used K, not °C or °F … because that’s what you have to use when analyzing heat engines.
The truth is, the climate system is an extraordinarily stable heat engine, and on my planet at least, the only way to explain such stability is the existence of a natural governor keeping it within range.
w.
PS—From a comment of mine above, here are the reasons you need to use K for thermodynamic analysis:
===
===
Richard,
Did you not catch the part where thermodynamics / heat engines only understand K (Kelvin). So when we are talking about the remarkable governance of our climate system, the percentage change in K is the only scientifically meaningful temperature statistic. Consider what we would get if we arbitrarily set your 14C 1975 temperature a new reference value of 0. Then your percent increase in temperature to 1 (formerly 15C) would be division by 0 and thus infinite or more properly, undefined.
The global warming since 1900 is real and is not +0.2%
Stow your thermodynamics where the sun don’t shine. I took a thermodynamics course in college on my way toa BS degree. But I try to explain climate in simple terms that ordinary people can understand without deliberately trying to deceive people with +0.2%, which you seem to celebrate.
Richard,
You patronize people when you simplify to the point of deception. You can’t “stow your thermodynamics” when you are talking about earth’s great heat engine. You did not refute what I said, you deflected. And why do you have to be such an arrogant jerk in replying to those who disagree with what you say?
The average temperature is a statistic, not a measurement
Ah, fine, you changed your mind and learned s.th. 😀
Three day ago you just told the opposite 😀
Hairy Krishna
Average temperature is a statistic based mainly on measurements. If I said anything else in the past 26 years, please show me where and I will add a correction. Until then, quit flapping your gums.
What ever you write now is BS as you wrote that averages are measured. What they are based on doesn’t interest here, they are the result of a calculation not of a measurement as you tried to explain 3 days ago.
Start with the 14°C, add 7%, equals 14.98°C
14°C = 57.2°F,
Start with 57.2°F, add 7%, equals 61.2°F = 16.2°C
Question: does the 7% warming result in a new temperature of
14.98°C or is it 16.2°C ?
Most average temperatures are reported in C. degrees and most people in the world use C. degrees.
99 percent of countries on Earth use Celsius to measure temperature, but the US doesn’t
C degrees is by far the best temperature scael to use when describing long term absolute temperature global warming over long periods of time
In the US people would be more comfortable hearing the global average temperature increased +2 degrees F. since the 1800
NASA claims the global absolute temperature is now about 59 degrees F
It has increased about 2 degrees from 57 degrees F. That is an increase of +3.5%
+7.0% would ne interpreted as small
+3.5% would be interpreted as small
+0.2% would be interpreted as near zero and would be a deception.
Stick your math gymnastics in the garbage can where they belong.
It has increased about 2 degrees from 57 degrees F. That is an increase of +3.5%
So 59F is an increase of 3.5% over 57F, but 13.89C to 15C (the same temperatures in a different scale) is a 7.5% increase. You post pretty much those numbers.
I’m going to assume that “ne” was meant to be “be”
Let’s double the amount of warming: 57F to 61F is 7%, which you said would be considered small. 13.89C to 16.11C is 14%. Would 14% still be considered small?
Richard, what a typical climate scientist you are …
Please tell me what is an increase from MINUS 5C to PLUS 5c in percent?
57F degrees is too cold to live outdoors for very long without some protection in the way of warm clothes and the like.
“That is an increase of about 7% since 1975”
And if you convert to F… you get a 3% change… that’s how DUMB your calculation is!
Again proving you have absolutely ZERO mathematical understanding of what you are doing.
You really are showing up your ignorance this time. !
And if you convert to K you only get a 0.35% change.
The Don Rickles of the comment section replies with his usual Forrest Gump wit.
As Willis explained, you should be using Kelvin instead of Celsius.
Willis is using K degrees to deceive people. If there was a lower percentage from C or F degrees, i suspect he would have used one of those scales
.
K is the scientific unit 😀
Ok, not for you as you are always far from science that deceives you 😀
Why do you insist on making things up? Willis has made it crystal clear why he used absolute temperature. In response to another comment he gave references for why it is the correct thing to do in the context of what he’s talking about.
Richard, that’s a damn lie. I’m an honest man, and I don’t “deceive people”. You’re either thinking of your friends or listening to the voices in your head.
The intro to any thermo text explains why all thermo calculations must, have to, need to, can only be in K.
You keep trying to do temperature percentage changes in °C or °F.
Suppose you tell us, then, what is the percentage change from 1°C to 2°C … 100%???
Really? A 1°C change in temperature is a 100% change, a doubling of the temperature? That’s crazy talk.
Sorry, but thermodynamics doesn’t work that way, for all the reasons I gave elsewhere.
w.
How to disable those leftist biased AI tools
Thousands of conservatives must contact the various AI’s and demand a calculation of the value of pi
That should keep those pesky AI confusers busy for a long time.
So you’re an old Star Trek fan?
How many times did Kirk beat a computer by asking it just the right question? 😎
Go one further and ask it the full and complete value of pi – ie 22/7.
I’d probably just ask it to play itself a game Tic-Tac-Toe.
Maybe it would learn that when it comes to Man trying control Nature, “Resistance is Futile”.
https://youtu.be/ijVijP-CDVI
As I understand the principal of “large language models” they basically determine the most statistically likely next word.
This means that they will reproduce the most frequently published narrative. The extent and deliberate nature of gatekeeping of the published scientific literature on climate was first exposed in 2009 ClimateGate email scandal, where we got to see exactly what all these “honest and objective” scientists were up to. It has not got any better since that time.
Distilling to most commonly repeated narrative is certainly NOT a valid means of checking a scientific question.
Maybe it could also tell us whether throwing tomato soup and classic masterpieces is likely to reduce the mean surface temperature.
No, it doesn’t work like that. The statistic is against its correlation to everything it was ever trained against. And using “Transformer” technology, those correlations are in a context too.
Its all very abstract but the upshot is that if the word car was statistically mostly followed by the word door in the training material, the way its trained doesn’t mean that whenever it mentions car, the next word is door let alone produce something that it was actually trained with.
Thank you Willis – my IQ always goes up a few points whenever I read one of your articles!
Here’s a silly idea – I’ve always wondered why the warming trend from about 1910ish to 1940 stopped and a slow decline occurred until about 1975. There was WW2 and the reconstruction, so CO2 way up (which I love to point out to alarmists) but also lots of spot and SO2, and such that would block sunlight very effectively. And you mentioned the atomic and fusion bomb testing in the South Pacific – that big, deep solar energy absorber!
Is it possible, like Tambora, Krakatoa and so on, that the A-bomb test could be a significant contributor to the cooling, or at least lack of warming (counting on which temperature series one looks at), that occurred during the 3 quarter of the 20th century?
Is it possible that the improved environmental standards that started in the early 70s, lead to less pollution in the air, thus allowing the clear warming trend in the 80s and 90s to occur (I’m thinking of UAH satellite temperature series – there’s a steady trend of the wiggly temperature graph until the El Nino of 1997/8 where the trend tends to take a 0.2°C step after each El Nino instead of there being a steady growth – but that’s my interpretation only, a straight trend fits the graph too but seems to be ignoring the details.)
Thanks again for a great article.
If step changes of ENSO conditions in an warm ocean phase might be responsible for an overall warming effect then why can’t step changes in ENSO conditions in a cool phase be responsible for cooling? Have a good look at the natural conditions prevailing at the time before going for something man-made – don’t make the same mistake made by climate enthusiasts of jumping at human-origin sources instead of natural sources.
The warming trend 1910 – 1940 was due to the immense amount of coal being burned and the soot but especially, the Sulphur that came with it.
Soot and Sulphur both Make Plants Grow Like Fook – it is why farmers in Germany used to fight each other over who could own land that a railway line ran through. The the soot, smoke, cinders and Sulphur were immense free fertiliser.
Plant life surged because of all that, plant death also and the dead plants, in the ground/soil/dirt retained water within the landscapes.
This tended to cool the summers but warm the winters – the overall effect being warming
From 1940 on (1945 strictly), plant death surged.
This being because allied governments did not want any repeat of food shortages and rationing.
Thus farmers were positively encouraged to use the vast amounts of nitrate that came available from unwanted munitions factories (and the tractors now pouring off production lines that previously built tanks) to plough plough plough therei land and grow wheat, corn, potatoes, beans, oilseeds etc etc
The notable thing being that these were/are ‘arable crops’
i.e. Annual plants
The endless ploughing destroyed the soil organic matter that had been built up over centuries from perennial grasslands where animals were kept as food – with only small amounts of arable land.
With the discernable of the soil organics, the water also disappeared so winters got colder and summers got warmer – overall a cooling
From 1975. Roundup appeared on the market and as far as soil organic matter was concerned, It Was The End. Finito. Kaput
Not very least as most soil organic matter in the form of bacteria and Glyphosate is an anti-biotic.
But for the mots part, it was used as a desiccant (shortening the active growing season) but also a a ‘poor/lazy man’s plough.
Soil organic matter plummeted, soil moisture content likewise and when that happened, the ‘Emergent Phenomena’ that depend on abundant water resource ceased to operate as it did.
This, for temperatures as recorded over ‘dry land’, was a myriad whammies of disaster.
No water for evaporative cooling
No water to make clouds
No water vapour to create rising air columns (thermals)
For emergent phenomena over the ocean, this was dream come true as it gave all the air the ocean had hoisted aloft making thunderstorms a place where it could easily & without impediment fall back to earth
And when it fell it heated.
So for the land, temps skyrocketed because of:
No evaporative cooling.
No clouds
Low albedo from all the ploughing and weed-killing
Low horizontal convection
Little rain
Not least the Coup-de-Grace, Foehn Effect (Katabatic) Heating from all the air descending after the ocean had hoisted it
BUT, the overall effect on Global Average Temp would be modest.
Yes the land heated, but because the T-storms over the ocean were now able to accelerate themselves, the oceans experienced a cooling effect – so tending to cancel out the warming over the land masses
And here we are now.
The significant problem is that that regime (accelerated ocean cooling) and high-pressure katabatic heating are self-reinforcing conditions.
Made worse beyond everyone’s worst nightmare because, as we’re endlessly told by both sides, the ocean is where Earth stores its heat
And what happens when the ocean runs out of heat if not = Ice Age
haha *discernable
Doncha just love artificial intelligence – make that ‘disappearance‘
German farmers were gutted when, circa 1923, Germany succeeded in electrifying nearly all its trains = no more smoky old steam engines belching all that free fertiliser
And yet, Peta, DB were still running some steam locos on the main line in the early 1970s whilst in the UK we removed them in 1968!
The Earth is still in a 2+ million-year ice age named the Quaternary Glaciation in an interglacial period that alternates with glacial periods. Over twenty percent of the land is frozen. https://en.wikipedia.org/wiki/Quaternary_glaciation
How about, “The major control of the earth’s temperature is the sun — but it moves so little and so slowly that nobody notices.”
Are humans doomed to extinction?
The question of human extinction is complex and involves many factors. Here are some perspectives based on various sources:
In conclusion, while there are threats and challenges, there are also factors that make humans resilient. The future of our species will likely depend on how we manage these challenges and adapt to changes. It’s a topic of ongoing research and debate among scientists.
Stupid thing doesn’t know about the dooming and prevaricates.
We don’t really know why some things — like sharks, brachiopod genus of lingula, coelacanth, birds, etc. — have survived mass extinctions, while others have relatively short lives. Until we better understand why some things survive mass extinctions, it is probably premature to speculate on how long humans will survive.
We just have to ask Schwab.
The only correct part of the term “artificial intelligence” is the “artificial” part as in an artifact created by people. Intelligence is not a feature of machines. They are simply designed well or badly to do what their designers intended.