Innovation, Hype & Failure: Elon Musk, Meta, Yuval Harari & AI (Part II)


Part I

One modern innovator is Elon Musk. Is he a modern-day Edison or a wealthy snake oil salesman? In 2013, Musk released a “Hyperloop Alpha Paper.” Smil is not impressed with Musk. He points out that “Hyperloop” is an incorrect and misleading term [111].

Musk, in July 2017, tweeted that he received verbal government approval for The Boring Company to build an underground NY-Phil-Balt-DC Hyperloop [121]. Smil points out that "any knowledgeable person would view this “with utter disbelief" [121].

With respect to innovation generally, Smil points out that people don’t appreciate “the need for indispensable plant nutrients in securing abundant and affordable food supply” [123]. In 1898, people were predicting that the world would run out of food by 1930 [124]. Today, “40% of the global population receive their dietary protein (directly from crops and indirectly from animal foodstuffs) from harvests that got nitrogen from the Haber-Bosch synthesis of ammonia [originated in 1910s]” [126].

Smil makes very good points about “techno-optimism, exaggerations and realistic expectations.” Misinformation gets traction because “in modern societies where masses of scientifically illiterate, and often surprisingly innumerate, citizens are exposed daily not just to overenthusiastically shared reports of potential breakthroughs but often to vastly exaggerated claims regarding new inventions” [152].

He talks about the colonization of Mars, brain-computer interface, and AI. With respect to Facebook’s prospect of living alternate lives as lifelike avatars in a realistic 3-D virtual space: “Of course, the most prominent testament to this delusion is Facebook’s 2021 conversion by renaming itself Meta and believing that people would prefer to live in an electronic metaverse (I cannot find suitable adjectives to describe this mode of reasoning, if that is the right noun to describe such an action)” [154].

Concerning AI: “But perhaps no category of modern inventions and technical advances has been so poorly and unhelpfully covered as AI” [157]. “The conclusion is obvious: our quest for AI is an enormously complex, multifaceted process whose progress must be measured across decades and generations” [159].

Smil makes a very interesting point regarding people’s perceptions of the pace and effectiveness of innovation. “Nothing has affected, and warped, modern thinking about the pace of invention and the extent of innovation than the rapid exponential advances of solid-state electronics” [160]. The famous Moore’s Law: A doubling every 18 months of computing power on a silicon chip—but this doesn’t apply to other fields. Regarding the various claims of exponential growth: “Just sit back and relax” [162].

Smil comments on the work of Yuval Harari and his notion of “Dataism” and that everything will be known and people will live forever. Interesting, but fanciful and detached from reality. An interesting bit of hype.

Smil’s request, after surveying invention and innovation, is that humanity should tackle some key problems: “Could we not come up with a manageable number—say, a score or two—of the most desirable items based on two overriding needs: to improve the fundamentals required for dignified life of the world’s population, and to do so without excessive impacts on the biosphere?” [171]

Some generic lessons regarding the pace of innovation (such as declaring a war on cancer): “Basic (scientific and technical) understanding must precede specific applications (perhaps the most obvious but repeatedly ignored reality); critical variables may get worse before they get better; it is unwise to specify outcomes by dates; every near-term target, no more than ten years away, will be missed” [176].

The EV mandates from governments are in his crosshairs. Proclamations are great but utterly improbable. Regarding EVs, Smil does the math! In 2021, there were 1.4 billion motor vehicles; only 17 million (1.2%) were electric and 99% were diesel or fuel. Having 40% of vehicles decarbonized by 2030 results in 570 million EVs [179].

He also points out some of the fallacies about promoting environmentally oriented things such as wind turbines. Smil notes that “the construction of massive wind turbines requires considerable quantities of reinforced concrete (cement and steel) for the foundations, steel for towers and nacelles, plastics for large blades, and lubricating oils for smooth motor service” and they all need to be delivered via transport vehicles [179].

Vaclav Smil is a unique commentator, combining the skills of a scientific historian and political commentator. This book, though technical in parts, is worth the read for the takeaways.