I LOVE this idea! The thing that bothers me most about Fossil Future is that it confirms all my biases. That's a real red flag for me. I can't possibly be that right about anything!

Expand full comment

Epstein does not challenge the basic premise that rising CO2 levels have increased global temperature, though he does challenge the more apocalyptic predictions. Rather, he argues that no matter whether earth gets warmer, cooler, or stays the same, there is no substitute for the use of fossil fuels to advance human flourishing and achieve "climate mastery" -- i.e., the ability to cope with climate problems no matter what they might be.

At root his argument is not about the science but about the moral premises we use to interpret the science -- specifically, about (1) the importance of a pro-human moral perspective and (2) the fallacy of focusing solely on the negative side-effects of fossil fuel use while ignoring their enormous life-or-death benefits.

Expand full comment

I've thought a lot about Epstein's book, but there are still many things that I wish I had more time to look into. Here are a few nominations:

1. Epstein waves away research on 100% renewable energy systems without any serious engagement. Maybe start with https://ieeexplore.ieee.org/document/9837910 and related sources.

2. The greening/fertilization effect is brought up frequently in Fossil Future. The impact on food systems and ultimately human well being is way more complex than "green goes up" and it would be nice to read a more measured analysis. Not sure where to start reading with this one.

3. Epstein talks exhaustively about the historical benefits of fossil fuels, but never talks about the economic orders and coercion that brought about such dominance. Would be interesting to contrast Malm's "Fossil Capital" (sorry, not exactly short) with the libertarian ideals of "Fossil Future".

I would warn that limiting yourself to "narrow" empirical issues may be inadequate. The climate system as a whole is too complex and interconnected. Epstein clearly uses this fact to curate topics that support his biases. Instead of responding only auditing topics found in Fossil Future, it's important to also look at Epstein's omissions (as per his own advice).

Expand full comment

Epstein says that climate models are excessively pessimistic.

These links argue that they have been accurate.



There is a further argument that models are underpredicting or failing to foresee some of the bad things happening now, indicating that the consequences of climate change are worse than these models state. Supposedly the “heat domes” seen in the West in 2021, which caused prolonged temperatures of over 120 F in some places, were not foreseen by any climate model. Last year was very bad for both droughts and flooding, depending on where you were. https://nypost.com/2022/08/28/pakistan-flooding-deaths-pass-1000-in-climate-catastrophe/

Relatedly, Epstein arguably understates challenges for fossil fuels. Morton Downey, Oil 101 (Wooden Table Press 2009) at 26, “Even in the most extreme optimistic scenario, conventional oil production will effectively cease to exist well before the end of this century – a fact even the most optimistic oil company agrees with.” Id. at 26. If that’s right, we’re going to have to replace fossil fuels no matter what.

Expand full comment
Jan 7·edited Jan 7

I think if you accept Epstein's philosophical framework, then there is only really two ways he could be wrong: Either he could underestimate the cost of unmitigated climate change, or he could overestimate the cost of switching to a renewable energy source. There seems to be plenty of people arguing that he does both.

1. The cost of switching to an energy system based on wind/solar.

Epstein claims that switching to a pure wind/solar energy system would be prohibitively expensive. Essentially based on the fact that no such energy system currently exist.

But there are plenty of people arguing that such a system is not only possible, but quite affordable. I would start with this review of 100% renewable energy systems: https://ieeexplore.ieee.org/document/9837910

2. The economic costs of unmitigated climate change

Epstein seems to dismiss most of the evidence for high climate costs, because he doesn't trust our "knowledge system" that produces it. But that excuse seems dubious, given how diverse the group of people, who warn us about these costs, really is.

I remember the Stern Review, which argued that the annual GDP loss could eventually be as high as 20%. https://en.m.wikipedia.org/wiki/Stern_Review

But there are also reports from Deloitte or SwissRe, who similarly project massive losses in GDP. https://www.deloitte.com/global/en/about/press-room/deloitte-research-reveals-inaction-on-climate-change-could-cost-the-world-economy-us-dollar-178-trillion-by-2070.html


A look at the wikipedia page for the economic impacts of climate change generally show quite scary numbers as well, or numbers that at the very least might be higher than the cost of switching energy system.


Expand full comment

I'm not sure if Epstein commented on this or not, as I haven't finished the book yet, but I have seen claims that "arable land" will be reduced, causing famines. I think the opposite is true.

Expand full comment

Nuclear is a dangerous/not-dangerous source compared to existing (and pie-in-the-sky) alternatives. I am convinced that it's the best option we have to address global warming. Is this wrong?

Expand full comment

Anybody would welcome an audit, if the auditor is up to the job. He's bound to have made mistakes in a 300+ page book, and unlike those in the alarmist community, will have more gratitude than chagrin that you've tried to improve the book.

Expand full comment

1) Chapter 5, the section titled "Ultra-Cost-Effective Mobile Energy"

Quotation: "Take cargo ships, the enormous machines that transport most international goods, which are 12 percent of mobile energy."

Claim: Cargo ships are 12 percent of mobile energy

Epstein's source says that all marine vessels in 2012 accounted for 12% of transportation energy consumption. Note that cargo ships are a subset of marine transportation. Marine transportation also included passenger vessels like ferries and cruise ships.

The distinction is important because passenger vessels have much shorter routes than cargo vessels and need to store less energy. There are some attempts to use alternative fuels for passenger vessels.

2) Chapter 5, the section titled "The Second Ingredient. . ."

Quotation: "Natural gas. . .is easy to transport cost-effectively via pipeline overland, but not so much over water (though this is improving with 'liquified natural gas'"

Problem: LNG shipping costs can be very low.

One source (https://www.insidelogistics.ca/opinions/lng-shipping-costs/) put the January 2020 cost of shipping LNG from the US to Japan at $1 / mmBtu. That is $3.41 / MWh.

Compared to the cost of electricity it generates, that is low. The October 2021 estimates for the cost of energy (https://www.lazard.com/media/451881/lazards-levelized-cost-of-energy-version-150-vf.pdf) put the cost of natural gas (labeled "gas combined cycle") at $45-74 / MWh in the US.

3) Chapter 5, the section titled "The Second Ingredient. . ."


Any competitor to silicon would have to overcome two hurdles.

One, silicon has remarkable natural properties as a semiconductor. . . Two, the modern microchip industry has spent generations figuring out specifically how to harness the properties of silicon in every kind of microchip imaginable, from computer microprocessors to the chips in smartphones to the chips in your car.

Problem: The semiconductor industry regularly uses materials other than silicon.

One of the first commercial semiconductors was lead sulfide used in crystal radios invented in 1906. A common, and earliest diode is the germanium diode.

About a dozen other semiconducting materials are used as alternatives to silicon but use the same industrial and technology base as silicon semiconductors. Semiconductor processing is not very specific to silicon. Growing silicon ingots is an exception.

Relevant to the power industry, SiC and GaN are used for high-power devices. The transistors used for high voltage DC transmission and many electric cars are SiC. GaN is growing in this market.

This is an important point because, throughout the chapter, Epstein ignores that much of the electrical power system is shared regardless of the power source. This is very analogous to how SiC devices can use the same technology and manufacturers as Si semiconductors.

In the electrical system, the entire transportation and distribution system is used regardless of the fuel source. That system is a component of consumer electrical costs. An important exception is on-site solar.

4) Chapter 5, the section titled "The Second Ingredient. . ." in the list of "four key elements of modern electricity's cost-effectiveness."

Quotation: "The leading fuel for peaking power is natural gas, which as a gas, is far either to rapidly adjust than a solid fuel like coal."

Problem: The real reason has nothing to do with the type of fuel and all to do with the plant design and economics. Peaking power plants are designed to provide peak power (energy per unit time) at a minimum cost. Baseload power plants are designed to provide energy at the lowest cost. This is a capital cost vs. operating cost tradeoff.

Currently, gas turbines are the most common peaking power plant because of the low capital cost to make a high-powered turbine plant. Before shale gas, they were oil powered turbines. In fact, many current turbines (like https://www.ge.com/gas-power/products/gas-turbines/7e) are designed to switch between natural gas and oil while running.

There are examples (like https://www.nrel.gov/docs/fy14osti/60575.pdf) where baseload coal plants were modified to be peaking plants with small physical modifications. This example uses a typical coal-fired boiler to run a steam turbine. It can turn on and off four times a day and lower its output from 480 MW to 90 MW. But, it takes over an hour to start. That is not as good as a gas turbine that can turn on and off within a minute.

It is true that coal is not a great fit for gas turbines. But, the problem is with the impurities in coal that do not burn and build up ash in the turbine ("Coal-fired gas turbine and ash separation system" https://doi.org/10.1016/0016-2361(95)00301-0). This is a problem with coal and not with it being a solid.

5) Chapter 5, the section titled "The Second Ingredient. . ." in the list of "four key elements of modern electricity's cost-effectiveness."

Quotation: "The steadiness [of peaking power] contributes to cost-effectiveness by using energy extremely efficiently."

Problem: Load-following and peaking power plants can be as efficient as baseload plants.

See problem #4 where the same baseload plant was modified to operate as a peaking plant. The changes made were completely unrelated to the efficiency of the plant.

Combined cycle plants are the most efficient plants. They are a combination of a gas turbine like on a jet aircraft and a steam turbine like on a coal plant that is heated by the gas turbine exhaust. They convert heat into rotational energy at up to 60% efficiency. They are used for both base load and peaking power plants. However, they are more commonly used for baseload and they are not as good at following fast changes as gas turbines.

Using a combined cycle plant for peaking power is like having a standard, peaking gas turbine connected with a steam turbine that is operated like a peaking plant.

The best reference I could find for peaking combined cycle plants is an article talking about how to make them better (https://doi.org/10.1016/j.applthermaleng.2016.12.004). I suspect this is a case where industry developed the capability so there are few academic citations.

Problems #4 and #5 appear to be caused by attributing economics and commercial agreements to technologies.

I think there are larger problems with omitted factual statements and analytic flaws. But, you asked for concrete facts.

Expand full comment

AE reports (Figure 2.1) "Hansen's predictions" from a newspaper, and finds them grossly inaccurate. However other reports (https://www.realclimate.org/index.php/archives/2018/06/30-years-after-hansens-testimony/) produce a very different assessment. AE is using Hansen 1986, not the far-more-famous-1988, and I kinda suspect that isn't an accident. So the question is: has AE fairly reported Hansen's work around that time? (I'm also doubtful that the newspaper he relies on reported accurately; I would fault AE for not going back to the original). See-also http://mustelid.blogspot.com/2023/01/lets-audit-alex-epstein.html

AE also asserts that the SPMs grossly distort the underlying IPCC reports. This is a large charge, and an important one for him, but he is desperately thin on detail.

Expand full comment

I confess that I haven't read the book, but a review in Foreign Policy led me to a Google Books quote from page 333: "The reality of today's climate knowledge is that climate scientists lack the causal understanding of climate needed to make meaningful predictions". As far as I can tell, he's saying that about climate science in general (!), in which case you should audit this in the context of the bet you made with me about global average temperature (about which BTW my 2023 update is coming soon!): https://standupeconomist.com/2022-update-on-my-global-warming-traffic-light-bet-with-bryan-caplan-and-alex-tabarrok/ (Additional relevant readings include my Cartoon Introduction to Climate Change, 2nd ed, which---despite apparently not having enough about public choice theory for your taste!---has some good evidence and links about meaningful predictions from climate science. https://standupeconomist.com/#writings You could also, of course try the latest IPCC report. But start with our bet... which should be "meaningful" in the same way that the Simon-Ehrlich debate was! Cheers, Yoram Bauman PhD

Expand full comment

I applaud your audit.

I know of one issue that I didn’t notice him covering. It may be so unrealistic that he skipped it, or covered it in a sentence of two and I didn’t make the connection. Unfortunately, I have now forgotten which doomsayer brought this to my attention. I only noticed it mentioned once even outside the context of Epstein's book, which seems to indicate it is not considered significant even by those advising climate panic.

The idea is that climate patterns and structures, wind patterns and trade winds are nonlinear, and the changes in CO2 or temperature might cause a sudden change rather than a gradual shift. The result could change rainfall patterns and local climates, making many of our staple crops and workhorse agricultural areas less viable or unviable. Some places might become flooded, while previously fertile agricultural land becomes desperate for water, not due to temperature directly, but because rainfall patterns shift in the extreme, out of proportion.

This doesn’t seem completely implausible, though I’m not sure if it is based on anything other than speculation and motivated reasoning. Sounds like the sort of thing Taleb would worry about. Unfortunately, that doesn’t seem to give any hint about where the danger is exactly, besides advising extreme and probably unwarranted caution. How fast is that steamroller going, and are those nickels or gold doubloons lying in front of it?

Expand full comment

I think that he underestimates what markets and people can achieve with a price on co2 but that is not so specific. I think that maybe enhanced weathering might be a cost effective way to reduce co2.

Expand full comment

The causal relationship of GDP to exclusively energy is a significant bias. Sanitation, governance, communication all have played a role.

Expand full comment

Land use and cost required to be fully reliant on renewables (solar & wind)

Expand full comment

I haven't read Epstein's book but I'd love to see an audit of the claims that wind and solar are the cheapest ways to generate electricity when all costs are accounted for - e.g. transmission, redundancy required for a grid with substantial intermittent sources, etc.

Expand full comment