Let’s Upgrade Bill Gates’ Climate Reading List

Where did he go wrong with this “miracle" fallacy?

The kernel of Gates’ mistake goes back to his reliance on “the top scientists.”(1) When Bill Gates tells you he’s sat with “the top scientists, and here’s what I’ve learned,” you just want to believe him. Bill Gates + top scientists = good enough for me, right?

Well, I pulled on that thread. A number of people like Jigar Shah, Andrew Shapiro and others have challenged Gates’ miracle fallacy. I don’t want to rehash their strong points, but rather to dig into the source code.

Gates has specifically referenced Vaclav Smil as one of his most respected energy advisors and writers. He has sung his praises repeatedly, and others are listening. Gates’ reading list is very popular, and now people like Mark Zuckerberg reference Smil as someone they turn to as well.

The conclusion Gates came to after reading Smil’s book Energy Myths and Realities was that we need miracles to the carbon challenges we face, because no technology solution in production today is energy-dense and/or cost-effective enough to scale in a timely fashion. This belief has been often repeated by Gates, most recently by him and his wife in their February 2016 annual letter from their foundation.

The problem with this thinking, and specifically the heavy reliance on Smil and his work, is that we do not need additional miracles. We have plenty of miracles already available. We need to scale these pre-existing technologies as quickly as possible. Thankfully, this is happening now, but it would happen even faster if Gates could change his rhetoric to match reality.

Vaclav Smil is clearly a smart man. However, his book, which appears to be the 10 commandments of clean energy skepticism for Gates, is filled with significant errors. Gates appears to accept these errors as truisms: “Smil ends by listing a number of lessons that come out of the mistaken predictions of the past, all of which I agree with.”(2)

Fortunately, now that the work is somewhat dated, it’s easy to show the errors in his ways. Let’s review.

Smil focuses his attacks on the future of innovation around a “trio of myths and misconceptions”(3), including:

1. The promise of electric vehicles
2. Energy storage
3. Decentralized “soft power” (his term for renewable power  -- presumably to qualify as “hard,” you need to light something on fire)

Electric vehicles

There is a fundamental fallacy undermining Smil’s logic. He appears to believe that simply because an innovation failed to meet its promise in the past, it will continue to fail. For EVs, his example is Amory Lovins’ Hypercar. Hypercar, an early pioneer in EV and materials design, failed as a business. However, Smil makes the mistake of using this example as the reason the next major EV startup will obviously fail. Unfortunately for him, Smil, writing in 2010, chose Elon Musk to ridicule, sarcastically noting that this entrepreneur “expects his business to be ‘a raging success...worth multiple billions of dollars.’” (4) The lesson here is that you can’t build a statistical case off of a single subject.

Smil seems more obsessed with poo-pooing the psychological hype machines than with really attempting to understand how the advances in battery technology have changed the game. “Because the list of its key promoters and ‘founding owners’ includes assorted Silicon Valley executives (leaving aside celebrities like Clooney), many people have begun to assume that in such techno-savvy hands the electric car is now bound to follow the trajectory of personal computers or mobile phones. They will be badly disappointed.” (5) Ask a Tesla Model S driver (or a Nissan Leaf driver, for that matter), if they are “badly disappointed.”

Batteries (for EV and otherwise)

When considering the lifespan or “cycle life” problem lithium ion batteries have, Smil once again ignores the immortal words of Wayne Gretzky and refuses to look at where the puck is going. On battery life, he states “The normal expectation is for two to three years of service.” (6) Maybe that was true six years ago when Smil was doing his research, but it’s clearly not true today.

Smil seems to have given up on the ability of incredible engineers, designers and entrepreneurs to collectively propel us into a cleaner, smarter, better future. When speaking about battery advancement, he laments, “I have not seen any proofs that the transformation of most of these new designs… will make cars much better than the best-performing models of today. My best guess is that something like Edison’s frustrating experience lies ahead, and not any Intel-like rate of advance.” (7) While the assertion sounds reasonable, we now know in hindsight it wasn’t a great guess. In fact, lithium ion has mirrored the extraordinary cost reduction curve of solar, despite not changing the fundamental chemistry.

Wind and solar

Smil saves his most consistent and dismissive slings for renewable energy generation (what he terms "soft energy"). In his view, those promoting renewables have “a misplaced faith in technical fixes as the best solution to the complex challenge of ensuring a global energy supply.” (8) He goes on to blame "soft energy’s" failure on its roots in the counterculture of the 1960s, not on any technical merit. Again, this is a paper written with 20 percent of the back of the book dedicated to scientific references. Yet his chapter dismissing clean energy concludes with a comparison of soft energy to Maoism.

Philosophizing aside, what’s most disconcerting is when a data-driven thesis is based on false data. In analyzing solar cost data, Smil describes a seemingly asymptotic decline in the costs of solar “cells”:

“Undoubtedly, PV cells have been getting cheaper. Modules cost more than $20 per peak watt in 1980, about $10 by 1985, and around $5 a decade later; but the price was still close to $4.50 at the end of 2009.” (9)

Giving him the benefit of the doubt here, Smil must have confused “PV cell” costs with “total installed cost per watt of a solar system.” In 2009, PV module prices were below $2/Wp and dropping fast. Of course PV cells, a component of a module, were much less in terms of cost per watt ($1.05/Wp)(10). More importantly, the cost reduction was not asymptotic. On the contrary, the costs plummeted further in the following years, and have dropped nearly another 75 percent (10). This is a significant error, since he would have a point if he were talking about cells at this $4.50 price level, and we’d all be the worse for it. However, since the truth is quite different, his data has served to confuse and mislead people like Bill Gates.

Smil again used incorrect (and this time, unattributed) data when referencing solar modules: “Moreover, the PV industry now aims at reducing the price of solar modules from about $4.5/W by the end of 2009 to $1.5-$2/W within a decade, a rate of price improvement far more sluggish than that conforming to Moore’s law.” (11)

Alas, this too is simply false. At the time of the writing of the book, I and many other entrepreneurs and solar executives were in fact announcing efforts to push below $1/Wp for fully installed systems, with solar modules at below $.50/Wp. In fact, we crushed that target, selling solar panels at $.50/Wp only seven years after Smil's book was published. Furthermore, as stated above, the costs of solar modules in 2009 were already below the $1.50/Wp amount referenced as his decade-out target.(12)

What to do?

The list of surprising and misguided conclusions continues in this book: energy density, transmission systems, the lack of value in distributed systems, etc. But you get the idea. Bill Gates built a number of core theses for his own work in energy around Smil’s analysis, and yet that analysis is fundamentally flawed. Here are my suggestions:

1. Upgrade the reading list. I hope anyone with specific recommendations for papers, books and reports for Mr. Gates to read regarding energy solutions and the timeline to carbon reduction will jump in here and offer upgrades to the reading list. We’re all on the same team here, and we know Mr. Gates’ intentions are good.
2. Smil, click undo. Why not update this book? The logic throughout was flawed, and yet you have incredible abilities as a historian and analyst. Use those talents to right this wrong.
3. Gates, click refresh. Mr. Gates, it’s not too late. As you think about how to build the Breakthrough Energy Coalition, let’s refresh your numbers and assumptions  --  particularly around solar, wind, storage and transmission. Let’s also update the headlines. It’s time to recognize that the only miracle we need is the collective will to scale, build and demand change. Assuming (2) doesn’t take place, please stop quoting Vaclav  --  and stop perpetuating this miracle fallacy. What we need is effort  -- by every human, not just miracle workers.

I welcome Dr. Smil, Mr. Gates and anyone else to join this conversation.