Until Citizens United is overturned, we will never have “citizens” to vote for. $ will lift up the candidates who we get to “vote” for. What a sham.
Overturning Citizens United needs to be the top priority, sadly, before anything else. Given the climate? Not sure we can make it. Signed, The once-an-eternal-optimist
In your haste, the climate change attribution you sight for the SoCal fires was wrong. It's 35% more likely, not 35 times. I hope that can be corrected. Otherwise, this & all your posts are spot on!
> All I’m saying is, the increasingly gloomy idea that there was no possible way we could every deal with climate because AI would soak up every new electron that sun and wind could ever provide, may not be quite as true as it seemed to some a week ago.
I think your analysis is flawed. More efficient AI models may only be a small hiccup in the strategy of some who would like to claim that we need more energy. In reality, it's absurd to assume that computer scientists and big tech companies will stop and accept a current stage of AI development as "good enough" and seek the most efficient solution. As even the comments by Trump and others point out, the advacement of AI is just an arms race and whenever more efficiency is found, more complexity will be sought even more vigorously by attempting to build even larger and more complex models.
All throughout time, industrial society has always expanded its overall energy consumption, and efficiency is just a way to make room for a quicker acceleration of advancing technology even more so that the absolute amount of energy used still increases. I think it's pretty clear that under the current state of technological development where we develop tech for the sake of tech, there will be no decrease in energy consumption as long as the technophiles have carte blanche to develop even more technology.
Moreover, when one frames the discussion along the lines of whether we can make do with less energy for AI, you are giving up to the conclusion of the technophiles that we need AI at all. We should NOT take the mainstream environmentalist approach that we should find the greenest ways to supply energy to the bottomless desire of the technophiles and stop there. We ALSO need to find ways to fight against the useless energy consumption as well, such as AI for example.
In short, it is ludicrous to think that any gain in efficiency will stop the desire of the capitalists and rich to soak up more energy. The next stage will simply be a new AI model that negates that efficiency in absolute value and provides the same flimsy excuse to provide help to the fossil fuel companies. The only way to truly fight this is to fight the desire for energy itself and have a reasonable and RESPECTFUL upper bound that we should use.
Spot on. Efficiencies that temporarily lower costs and temporarily disrupt tech markets simply result in more production & consumption of whatever it is, per the Jevons Paradox. I challenge anyone to give an example that disproves the case. But it's not surprising to me that Bill skirts the issue by dwelling on the immediate effects of DeepSeek innovation, offering this as a Smidgen of Good News. After all, these days, he routinely exalts solar, wind & EVs as efficient alternatives to FFs even though they depend entirely on FFs in the making and cause further deep harm to Nature. Not to mention that "renewables" cannot be shown to have (or will ever have) the effect of reducing net energy consumption and net reduction of CO2e emissions. Noting that, well, we can't be sure how it will all shake out... is a caveat that rings hollow. We >are< sure about where it's all going; skyrocketing FF use and ever increasing energy consumption.
So, I have to disagree that desire for energy will or can be reduced. A reasonable and RESPECTFUL upper bound is not in the cards. There is no business or political advantage in the concept and no way to enforce it.
Perhaps, but I have to say of revolution what I say about degrowth and the supposed transition to a clean energy future – left undefined & with no plausible strategy or plan of action put forward, it's a boat that never leaves the dock.
Maybe, maybe not. The thing about revolution is that it doesn't require definition. When it happens, it happens. It's like COVID in that way: that was completely unexpected, and many will say the same of the decline of modern civilization.
Your mind only thinks of "plausible strategies" due to the conditioning of the machine. That's the only way we can think now, because we are entrenched in the system. But like a healthy body that cannot envision a freak accident, modern industrial civilization will face its demise, which from nature's point of view, will be like a welcome eradication of a fatal disease. Just as in humans we sometimes see miraculous recoveries from near certain death, so will the planet recover from us.
Moshe, on the merry-go-round of academic energy analysis there is no brass ring for reduction of total global emissions from the burning of fossil fuels. But that is the goal is it not? The goal is not arriving at a moderately bigger share of the power sector for wind, solar,etc. compared to FF generated electricity – as suggested by the study you cite, and by many others. The power sector occupies an historic and current ~20% of total energy. The other ~80% is from FFs. This study speculates that 67% of electricity could be generated from renewables by 2050, using (absurdly) 2005 as the emissions baseline... but only if major advances can be made in technological efficiencies, investment, adoption, regulation and science-informed policy... factors that are famously unpredictable.
Suppose this unlikely what-if scenario panned out. You would still have 33% of electricity generated with gas, coal & oil in 2050. And you would still have the other 80% of total energy provided by gas, coal & oil, spewing ever more billions of tons of CO2e. Because that is what's required for continued growth of the global industrial economy. It doesn't matter what I "want to believe". What matters is evidence. What matters is evaluating the prospects for decarbonizing energy use in light of ALL compelling and constraining factors. Analyzing the power sector in a silo, in isolation from the rest of the global energy system, only points up the flaws in doing so.
In this paper there is no acknowledgement of the 20%–80% ratio with FFs; no mention that the rate of FF emissions is rising exponentially to meet rising demand, overwhelming the capacity of renewables to have significant decarbonizing effect; no mention that the policy of all FF producing countries is to capture and use FFs without limit and without care for the consequences.
Also not included in this 300 page exercise in energy system myopia are the "externalities" depenably ignored in the tech–focused bubble: The complete lifecycle dependance of "renewables" on FFs, from mining to manufacture to their replacement in 10 to 30 years. The carbon footprint of each technology. The increasing scarcity of finite resources leading to elevated resource conflicts and higher costs. And the most troubling fact of creating & deploying more & more of these short–lived products: more & more damage to critical ecosystems from mining, watershed pollution, deforestation and road building. All of which is accomplished by using millions of barrels of diesel fuels to power gigantic scale excavators, trucks, processing/manufacturing plants and shipping–dependant supply chains.
I take the time to say all this because I think it's worth the trouble. I am not the one here who is ignoring "available research". I think that burden is on you – and on so many others – who take the false promise of renewables as a given without actually looking at the proposition with a skeptical (scientific) eye. So, best to you. But please... do your homework. It's complicated.
I would agree that adaptation needs to be the main if unrealizable focus. But perhaps we need to go full speed ahead with both an attempt at a zero carbon world and adaptation to the dire future. We can’t foresee everything. Perhaps humility along with a multiplicity of actions will be the best. Holding two opposed ideas and performing opposed actions seems the right path to me.
But that's the thing – there IS no full speed ahead attempt at a zero carbon world on anyone's or any institution's or any government's part. There is only the full-speed-ahead limitless growth project of capitalism conducted by the 10% of people who literally own all the levers of political & economic power. "We" have zero influence, however much you would like to think so. It is the hardest pill to swallow because it means we are fucked, at the mercy of a system that has already gone off the cliff. As for adaptation, government focus is elsewhere, occupied with preserving the status quo. No help there. "We" are on our own with vanishingly few options.
I live in California and don’t feel that way. My government is very much aligned with acting to mitigate climate change. But it is difficult and much of what we try has unforeseen negative tradeoffs. That doesn’t mean hopelessness is the preferred option or smart choice.
Indeed, something I am keenly aware of. And Jevon's paradox is a side-effect of the larger effects of technological determinism, first described by the Greeks in their conception of techne and logos, expanded later by Heidegger, Ellul, and Skrbina.
Yes, thank you Bill. And in your next note can you help us understand how we increase solar without lithium battery storage? Lithium is a filthy mining process destroying precious water that is now skyrocketing. See Saudi investments in it. It is solar’s ugly shadow. Signed, She-who-no-longer-sees- solar-as-our-savior.
Many battery technologies are available and getting better and cheaper.
Sodium-Ion batteries are especially great for grid storage. (China's CATL battery manufacturer is on their fourth or fifth edition of sodium-ion batteries in cars, scooters, and more.) Salt is a bit heavier than lithium, but the sixth most plentiful mineral on the planet that doesn't overheat and is more plentiful and safely/easily mined. (Not that there are no environmental impacts, but much less!)
To see the many varieties of batteries, I recommend a YouTube program: Just Have A Think. Fantastic, smart, balanced factual information and analysis.
Also, Stanford University Engineer Mark Jacobson is coming out this month with a new edition of his terrific "solutions" book:
Still No Miracles Needed: How Today's Technology Can Save Our Climate and Clean Our Air
We're in the midst of exciting "disruption curves". The fossil fuel industry, and nuclear power, can't compete with solar and wind plus batteries. We don't have to convince everyone to sacrifice or pay more for energy. Our challenge is to get as many people as possible to:
1) Divest from risky fossil fuel and nuclear technologies
2) Join in calling for an end to subsidies for energy sources that contaminate our water, air, and land and damage our health, cause cancer, cause death from air pollution.
3) Take advantage of cheaper, cleaner, safer, renewable solar and wind energy and reject dirty, limited, toxic, heavily subsidized, dangerous forms of energy.
We have "money-saving" on the side of solar and wind and batteries. Let's tell everyone we can! Every dollar shifted away from fossil fuel and nuclear industries is another step of hope.
The biggest miracle will be convincing other countries like Brazil, with extreme amounts of protectionism for their industries, to actually follow any of these steps.
Makes you wonder whether crypto and AI weren't both schemes by the energy industry. Crypto has no intrinsic value, and AI only for limited purposes. (Marketing images and text does not count.)
I can’t begin to tell you how grateful I am to read your excellent and factual writing on the climate and environment. If only every person would read it and do their part to stop the use of fuels.
Thanks for your comments on China's advances. Here are some thoughts about their vast deployment of rooftop solar:
In 2024 China installed 277 Gigawatts (GW) of solar, more than half of which was rooftop rather than centralized ground mounted. For comparison, California has about 80 GW of total installed power capacity. In other words, in 2024, China installed more rooftop solar power (around 140 GW) than California provides from all power sources.
How did they do that?
Mostly by unleashing fierce market competition, but also with government policy guidance.
With these forces, solar panels made in China have become incredibly cheap–now around 15¢ per watt. By comparison, when we installed solar on our roof in 2010, our Chinese-made panels cost around $3.00 per watt.
So solar has become much more affordable. But it’s not just the price. Rooftop solar has only become dominant in China in the past three years. Prior to that, the Chinese were mostly building large solar farms in western China, requiring expensive transmission lines.
But starting in 2021, a new plan to incentivize rooftop solar in China took off and has been wildly successful. The new plan ensures that all three players make money–the utility, the solar installation company, and the home/property owner. And there is one added ingredient–a local organizer.
The local organizer takes a neighborhood, or a whole village, and talks to all the property owners who have promising sites (i.e. no trees, tall neighboring buildings, or other obstacles to solar). These promising sites are all assembled into one big project, which allows the solar installation company to take advantage of the economies of scale.
Why couldn’t we do the same thing in the US? In the Bay Area we have Community Choice energy providers (CCA’s), which should be responsive to demands for more solar. If community organizers can team up with the CCA’s and the solar installers, we should be able to duplicate China’s success. Electrons behave the same way in China as they do in the U.S.
Alongside of efficiency gain, practicing energy conservation as well as promoting the "vision of a new economics (https://centerforneweconomics.org/about/our-mission/) could counter balance the rising of net energy consumption. A Circular Economy would also be a big player in reducing energy consumption.
Bill, I get that one AI model might be better than another model, but need to ask the question, are we setting the terms of the public debate on data centers? Thanks for all you do!
Good analysis. As is mentioned the "Jevons paradox" is around the corner whatever we do to reduce the use of energy/material etc. We are a long way from understanding the necessity to slow down and cut back on everything that is not essential for maintaining life. There is enough space for 8 billion people in a sustainable fashion which is not currently the case.
Great writing as usual! Michael Thomas, in his Jan 21st article on Substack, muses about the excess energy needed to run AI and comes to the conclusion that it's not very likely. DeepSeek, and the Chinese save the day - again (when it comes to clean energy and more efficient use of it).
Until Citizens United is overturned, we will never have “citizens” to vote for. $ will lift up the candidates who we get to “vote” for. What a sham.
Overturning Citizens United needs to be the top priority, sadly, before anything else. Given the climate? Not sure we can make it. Signed, The once-an-eternal-optimist
In your haste, the climate change attribution you sight for the SoCal fires was wrong. It's 35% more likely, not 35 times. I hope that can be corrected. Otherwise, this & all your posts are spot on!
I see the correction has been made! Congratulations to you for spotting and then acting to alert the author, and the author for making the correction.
> All I’m saying is, the increasingly gloomy idea that there was no possible way we could every deal with climate because AI would soak up every new electron that sun and wind could ever provide, may not be quite as true as it seemed to some a week ago.
I think your analysis is flawed. More efficient AI models may only be a small hiccup in the strategy of some who would like to claim that we need more energy. In reality, it's absurd to assume that computer scientists and big tech companies will stop and accept a current stage of AI development as "good enough" and seek the most efficient solution. As even the comments by Trump and others point out, the advacement of AI is just an arms race and whenever more efficiency is found, more complexity will be sought even more vigorously by attempting to build even larger and more complex models.
All throughout time, industrial society has always expanded its overall energy consumption, and efficiency is just a way to make room for a quicker acceleration of advancing technology even more so that the absolute amount of energy used still increases. I think it's pretty clear that under the current state of technological development where we develop tech for the sake of tech, there will be no decrease in energy consumption as long as the technophiles have carte blanche to develop even more technology.
Moreover, when one frames the discussion along the lines of whether we can make do with less energy for AI, you are giving up to the conclusion of the technophiles that we need AI at all. We should NOT take the mainstream environmentalist approach that we should find the greenest ways to supply energy to the bottomless desire of the technophiles and stop there. We ALSO need to find ways to fight against the useless energy consumption as well, such as AI for example.
In short, it is ludicrous to think that any gain in efficiency will stop the desire of the capitalists and rich to soak up more energy. The next stage will simply be a new AI model that negates that efficiency in absolute value and provides the same flimsy excuse to provide help to the fossil fuel companies. The only way to truly fight this is to fight the desire for energy itself and have a reasonable and RESPECTFUL upper bound that we should use.
Spot on. Efficiencies that temporarily lower costs and temporarily disrupt tech markets simply result in more production & consumption of whatever it is, per the Jevons Paradox. I challenge anyone to give an example that disproves the case. But it's not surprising to me that Bill skirts the issue by dwelling on the immediate effects of DeepSeek innovation, offering this as a Smidgen of Good News. After all, these days, he routinely exalts solar, wind & EVs as efficient alternatives to FFs even though they depend entirely on FFs in the making and cause further deep harm to Nature. Not to mention that "renewables" cannot be shown to have (or will ever have) the effect of reducing net energy consumption and net reduction of CO2e emissions. Noting that, well, we can't be sure how it will all shake out... is a caveat that rings hollow. We >are< sure about where it's all going; skyrocketing FF use and ever increasing energy consumption.
So, I have to disagree that desire for energy will or can be reduced. A reasonable and RESPECTFUL upper bound is not in the cards. There is no business or political advantage in the concept and no way to enforce it.
> There is no business or political advantage in the concept and no way to enforce it.
Revolution against the system is a way to enforce it.
Perhaps, but I have to say of revolution what I say about degrowth and the supposed transition to a clean energy future – left undefined & with no plausible strategy or plan of action put forward, it's a boat that never leaves the dock.
Maybe, maybe not. The thing about revolution is that it doesn't require definition. When it happens, it happens. It's like COVID in that way: that was completely unexpected, and many will say the same of the decline of modern civilization.
Your mind only thinks of "plausible strategies" due to the conditioning of the machine. That's the only way we can think now, because we are entrenched in the system. But like a healthy body that cannot envision a freak accident, modern industrial civilization will face its demise, which from nature's point of view, will be like a welcome eradication of a fatal disease. Just as in humans we sometimes see miraculous recoveries from near certain death, so will the planet recover from us.
Moshe, on the merry-go-round of academic energy analysis there is no brass ring for reduction of total global emissions from the burning of fossil fuels. But that is the goal is it not? The goal is not arriving at a moderately bigger share of the power sector for wind, solar,etc. compared to FF generated electricity – as suggested by the study you cite, and by many others. The power sector occupies an historic and current ~20% of total energy. The other ~80% is from FFs. This study speculates that 67% of electricity could be generated from renewables by 2050, using (absurdly) 2005 as the emissions baseline... but only if major advances can be made in technological efficiencies, investment, adoption, regulation and science-informed policy... factors that are famously unpredictable.
Suppose this unlikely what-if scenario panned out. You would still have 33% of electricity generated with gas, coal & oil in 2050. And you would still have the other 80% of total energy provided by gas, coal & oil, spewing ever more billions of tons of CO2e. Because that is what's required for continued growth of the global industrial economy. It doesn't matter what I "want to believe". What matters is evidence. What matters is evaluating the prospects for decarbonizing energy use in light of ALL compelling and constraining factors. Analyzing the power sector in a silo, in isolation from the rest of the global energy system, only points up the flaws in doing so.
In this paper there is no acknowledgement of the 20%–80% ratio with FFs; no mention that the rate of FF emissions is rising exponentially to meet rising demand, overwhelming the capacity of renewables to have significant decarbonizing effect; no mention that the policy of all FF producing countries is to capture and use FFs without limit and without care for the consequences.
Also not included in this 300 page exercise in energy system myopia are the "externalities" depenably ignored in the tech–focused bubble: The complete lifecycle dependance of "renewables" on FFs, from mining to manufacture to their replacement in 10 to 30 years. The carbon footprint of each technology. The increasing scarcity of finite resources leading to elevated resource conflicts and higher costs. And the most troubling fact of creating & deploying more & more of these short–lived products: more & more damage to critical ecosystems from mining, watershed pollution, deforestation and road building. All of which is accomplished by using millions of barrels of diesel fuels to power gigantic scale excavators, trucks, processing/manufacturing plants and shipping–dependant supply chains.
I take the time to say all this because I think it's worth the trouble. I am not the one here who is ignoring "available research". I think that burden is on you – and on so many others – who take the false promise of renewables as a given without actually looking at the proposition with a skeptical (scientific) eye. So, best to you. But please... do your homework. It's complicated.
I would agree that adaptation needs to be the main if unrealizable focus. But perhaps we need to go full speed ahead with both an attempt at a zero carbon world and adaptation to the dire future. We can’t foresee everything. Perhaps humility along with a multiplicity of actions will be the best. Holding two opposed ideas and performing opposed actions seems the right path to me.
But that's the thing – there IS no full speed ahead attempt at a zero carbon world on anyone's or any institution's or any government's part. There is only the full-speed-ahead limitless growth project of capitalism conducted by the 10% of people who literally own all the levers of political & economic power. "We" have zero influence, however much you would like to think so. It is the hardest pill to swallow because it means we are fucked, at the mercy of a system that has already gone off the cliff. As for adaptation, government focus is elsewhere, occupied with preserving the status quo. No help there. "We" are on our own with vanishingly few options.
I live in California and don’t feel that way. My government is very much aligned with acting to mitigate climate change. But it is difficult and much of what we try has unforeseen negative tradeoffs. That doesn’t mean hopelessness is the preferred option or smart choice.
What you are describing is a form of Jevons paradox.
Indeed, something I am keenly aware of. And Jevon's paradox is a side-effect of the larger effects of technological determinism, first described by the Greeks in their conception of techne and logos, expanded later by Heidegger, Ellul, and Skrbina.
Yes, thank you Bill. And in your next note can you help us understand how we increase solar without lithium battery storage? Lithium is a filthy mining process destroying precious water that is now skyrocketing. See Saudi investments in it. It is solar’s ugly shadow. Signed, She-who-no-longer-sees- solar-as-our-savior.
Hi, Lauren,
Many battery technologies are available and getting better and cheaper.
Sodium-Ion batteries are especially great for grid storage. (China's CATL battery manufacturer is on their fourth or fifth edition of sodium-ion batteries in cars, scooters, and more.) Salt is a bit heavier than lithium, but the sixth most plentiful mineral on the planet that doesn't overheat and is more plentiful and safely/easily mined. (Not that there are no environmental impacts, but much less!)
To see the many varieties of batteries, I recommend a YouTube program: Just Have A Think. Fantastic, smart, balanced factual information and analysis.
Also, Stanford University Engineer Mark Jacobson is coming out this month with a new edition of his terrific "solutions" book:
Still No Miracles Needed: How Today's Technology Can Save Our Climate and Clean Our Air
https://web.stanford.edu/group/efmh/jacobson/WWSStillNMN/StillNMN.html
We're in the midst of exciting "disruption curves". The fossil fuel industry, and nuclear power, can't compete with solar and wind plus batteries. We don't have to convince everyone to sacrifice or pay more for energy. Our challenge is to get as many people as possible to:
1) Divest from risky fossil fuel and nuclear technologies
2) Join in calling for an end to subsidies for energy sources that contaminate our water, air, and land and damage our health, cause cancer, cause death from air pollution.
3) Take advantage of cheaper, cleaner, safer, renewable solar and wind energy and reject dirty, limited, toxic, heavily subsidized, dangerous forms of energy.
We have "money-saving" on the side of solar and wind and batteries. Let's tell everyone we can! Every dollar shifted away from fossil fuel and nuclear industries is another step of hope.
The biggest miracle will be convincing other countries like Brazil, with extreme amounts of protectionism for their industries, to actually follow any of these steps.
Oops! Meant to write: There are environmental impacts to salt mining, but much less than lithium.
Bill, I just want to thank you for your continued effort to give us such in-depth climate news on a consistent basis. You are much appreciated!
Among the energy stocks down are uranium miners. Might not be the demand for nuclear power with DeepSeek. https://www.marketindex.com.au/news/uranium-stocks-tumble-as-deepseek-raises-doubts-about-ai-energy-demand
Makes you wonder whether crypto and AI weren't both schemes by the energy industry. Crypto has no intrinsic value, and AI only for limited purposes. (Marketing images and text does not count.)
Crypto is very handy for laundering money.
This isn’t the issue. The destruction of our civil service is.
I can’t begin to tell you how grateful I am to read your excellent and factual writing on the climate and environment. If only every person would read it and do their part to stop the use of fuels.
Elon and his shitty billionaire buddies can’t steal anymore money from the US government regarding AI. 👏🏆⚖️
Thanks for your comments on China's advances. Here are some thoughts about their vast deployment of rooftop solar:
In 2024 China installed 277 Gigawatts (GW) of solar, more than half of which was rooftop rather than centralized ground mounted. For comparison, California has about 80 GW of total installed power capacity. In other words, in 2024, China installed more rooftop solar power (around 140 GW) than California provides from all power sources.
How did they do that?
Mostly by unleashing fierce market competition, but also with government policy guidance.
With these forces, solar panels made in China have become incredibly cheap–now around 15¢ per watt. By comparison, when we installed solar on our roof in 2010, our Chinese-made panels cost around $3.00 per watt.
So solar has become much more affordable. But it’s not just the price. Rooftop solar has only become dominant in China in the past three years. Prior to that, the Chinese were mostly building large solar farms in western China, requiring expensive transmission lines.
But starting in 2021, a new plan to incentivize rooftop solar in China took off and has been wildly successful. The new plan ensures that all three players make money–the utility, the solar installation company, and the home/property owner. And there is one added ingredient–a local organizer.
The local organizer takes a neighborhood, or a whole village, and talks to all the property owners who have promising sites (i.e. no trees, tall neighboring buildings, or other obstacles to solar). These promising sites are all assembled into one big project, which allows the solar installation company to take advantage of the economies of scale.
Why couldn’t we do the same thing in the US? In the Bay Area we have Community Choice energy providers (CCA’s), which should be responsive to demands for more solar. If community organizers can team up with the CCA’s and the solar installers, we should be able to duplicate China’s success. Electrons behave the same way in China as they do in the U.S.
Alongside of efficiency gain, practicing energy conservation as well as promoting the "vision of a new economics (https://centerforneweconomics.org/about/our-mission/) could counter balance the rising of net energy consumption. A Circular Economy would also be a big player in reducing energy consumption.
Bill, I get that one AI model might be better than another model, but need to ask the question, are we setting the terms of the public debate on data centers? Thanks for all you do!
Here's my take... https://skywaterearthhobie1.substack.com/p/framing-data-centers
Hobie
Good analysis. As is mentioned the "Jevons paradox" is around the corner whatever we do to reduce the use of energy/material etc. We are a long way from understanding the necessity to slow down and cut back on everything that is not essential for maintaining life. There is enough space for 8 billion people in a sustainable fashion which is not currently the case.
So thorough on every front. Thank you so much.
Hi Bill
Great writing as usual! Michael Thomas, in his Jan 21st article on Substack, muses about the excess energy needed to run AI and comes to the conclusion that it's not very likely. DeepSeek, and the Chinese save the day - again (when it comes to clean energy and more efficient use of it).