Friday, June 14, 2019

Lessons in Counterintuition

1. A popular survey question purports to expose our innate irrationality. It goes like this: you can have 100,000 USD if your enemy or your ex-spouse gets 1 million USD. Apparently, most respondents declined the offer. 

But all one has to do is add more nuance to the scenario to get a different overall response, thereby exposing the original question as meaningless. Try this: you have non-dischargeable debt of 50,000 USD. You can have 55,000 USD if your enemy receives 550,000 USD. Answers to the second question will be more varied, indicating short-answer surveys don't offer enough nuance to justify their cost or relevance. 

When such deficient "research" is passed off as newsworthy, serious journalism has died in America, but I'm also worried about inattention to the sociology field. Medical advances, especially in neuroscience, are leading governments and academics to focus on psychology and pharmacology departments without the involvement of independent entities capable of institutional knowledge. Absent relevant and reliable anecdotal evidence, scientific researchers may spend taxpayer and other funds chasing chimeras. 

2. Speaking of a failure to appreciate nuance, people are worried about AI's ability to increase unemployment. 
One person believes AI may wipe out 47% of existing jobs in America,
an astoundingly specific number.
Yet, the AI problem may be even bigger than unemployment if the world's technological AI race gives existing leaders--not necessarily in power by merit--the potential to cement their advantage over others, snuffing out change from local sources. 
Roberto Unger's Free Trade Reimagined (2007).
Imagine a robot that can scan all local residents for weapons as well as criminal records. No longer would a rural recruit dropped in a foreign land need be in a position to kill unarmed civilians. No longer would a wary security guard at a private establishment need assume every patron a potential threat. 

But let's fast-forward to the future. The aforementioned technology has made it easier to invade and occupy different lands if only to prevent another competitor from doing the same. Thus, while such technology would make the weak and unarmed safer in the short-term, the long-term picture is unclear. Nevertheless, if modern history is any indication, one can imagine this technology leading to more occupation, then removal of armed resistance to foreign culture, and finally the supplanting of local culture, beliefs, and methods. In one fell swoop, the same AI technology that protected the weak and unarmed has now extinguished the capacity for the same residents to achieve Roberto M. Unger's "diversity" component--leading to perpetual dependency on a foreign power. 
From Unger's Free Trade Reimagined (2007). 
Unger discusses diversity in ways unlike any other economist or political thinker. In order for workers not to be left out as innovation and creative destruction are financed by larger players, he argues it is imperative that 1) local entities are able to innovate in their own ways, unconstrained by centralized norms (another way of saying local culture ought to be supported through "collective experimentation" rather than subservience to centralized market forces); and 2) all entities are able to disregard prior norms if doing so would improve conditions for both capital and labor. 
From Roberto Unger's Free Trade Reimagined (2007)
Unger's "economic diversity" is the characteristic most under threat with advanced AI--despite not a single politician articulating this potential problem apart from anti-trust concerns. 

3. More lessons in counter-intuition: Country A has an 80% poverty rate. Country B has a 50% poverty rate and a democratic political system. Without knowing more, which country has the better chance of avoiding societal cohesion problems in the next 50 years? 

You'd think it would be the country with less poverty, but America in 2019 proves that when at least half of a country is able to structure the tax code, government funding, and housing inflation in ways that benefit existing interest groups, anyone outside those groups is left behind not just relatively but absolutely. (For the economics wonks: I use these two terms informally, but Unger uses David Ricardo's comparative advantage vs. Adam Smith's absolute advantage as an overall framework, at the same time casting doubt on Ricardo's ideas due to their limited scope, i.e., trade between just two countries using just two popular products.) 

Where economic theory typically fails is its inability to properly incorporate the social costs of underinvestment, meaning over time, absent some mechanism--such as widespread and cost-effective public transportation, genuinely merit-based and affordable colleges, etc.--segregation occurs, cementing physical and abstract (e.g., communication) gaps and reducing opportunities for reconciliation. Worse, as existing winners gain more affluence, they begin to see others outside their increasingly closed-loop system as morally deficient, eventually rejecting public institutions as the costs of reconciliation increase exponentially every year effective solutions are not implemented. (American acceptance of exorbitantly expensive private K-12 schooling is one example of such a breakdown--as if even one K-12 school not properly educating future voters in a democratic system providing equal votes to each citizen is acceptable. As I've written before, "Generally, long-term costs of exclusion, even if unintentional, far exceed the costs of inclusion on the front end.")  

In contrast, a country with an 80% poverty rate cannot easily segregate the country excessively or irreversibly. Any national public works program must consider more rather than fewer residents by demographic default. Moreover, the cost of essential items such as housing cannot be inflated beyond a point of no return even with the assistance of the banking sector unless wages also rise among a greater percentage of the population. To sum up, it is better for individuals to be rich than to be poor, but not necessarily for countries. 

© Matthew Rafat (2019)

Bonus I: "Free trade will flourish when the rules of the world trading system are designed to reconcile openness and diversity, not to suppress diversity in the name of openness." -- Roberto M. Unger 

Bonus II: "Humanity can become more unified only by seeking to develop in different directions... [so as] to establish a machine for the creation of collective difference [that supports] alternatives by making the world safer for them." -- Roberto M. Unger 

Bonus III: continuing the third example above, one can see developing countries' biggest problem is not technological access, but corruption. Why? Because developed countries' need for more consumers, including ones willing to spend beyond their means, will lubricate technological transfers so as to establish platforms. A governor or president who chooses the wrong transportation company contract or who builds asphalt roads instead of Tokyo-style trains is a developing country's greatest threat to long-term success. 

Even if a developing country chooses well, only half of the battle has been decided--for example, if a train is chosen but goes over estimated costs, not only will the government lose taxpayers' money while further mortgaging its citizens' futures to foreign banks, it will also lose credibility, weakening its ability to govern and to regulate. 

Thursday, June 13, 2019

Tesla's Annual Shareholder Meeting (2019)

I won't spend too much time on Tesla's 2019 annual shareholder meeting, because it was a well-oiled (pun intended) marketing job, and as a great writer once said, "All marketers are liars." 
Having attended Apple shareholder meetings when Steve Jobs was at the helm, I'm familiar with the cult of personality, which often arises when an individual, against all odds, goes his own way. Like Elon Musk against Big Oil--one of America's linchpins in its military-industrial complex--Jobs was alone in making iOS more of a closed-loop system than a less secure, open-ended, Android one. Unlike Jobs, however, Musk has no charisma (perhaps due to Asperger's) and does not seem to view the supply chain as vital to innovation. Ironically, according to Dan'l Lewin, CEO of the Computer History Museum--where the meeting was held--Jobs' "focus on supply chain and inventory and those things was phenomenal." Unfortunately, no one would say the same about Musk, a deficiency that will surely allow competitors to catch up

Musk's lack of organization--a common trait in highly-performing individuals without disciplined support teams--was such that he missed his own entrance at the meeting, forcing the emcee to walk through a door to get him. Musk's obvious idealism and intelligence have earned him the benefit of the doubt; indeed, it is because of Musk that traditional combustion engine companies have been forced to play catch-up, no longer able to argue EV consumer demand fails to justify major investment. And yet, despite all of Tesla's positive points, Musk's work in SpaceX and with satellites are the most innovative--so of course no one asked him how the 1967 Outer Space Treaty should be updated, or what obstacles private companies faced in space exploration when most satellites are still government-owned. (I didn't get a chance to ask a question because too many shareholders representing third parties (VC funds? Marketing firms?) decided to ask softball questions.) 

I'll summarize Musk's most interesting comments below: 

1. The Model 3 is the best-selling car by revenue, with second place belonging to the Toyota Camry. 10 years ago, no one would have believed it. [Note: I wasn't impressed. "Best-seller by revenue" is a made-up metric. It was originally designed to convince corporations to invest in lithium battery technology for cars, a much higher return on capital than just laptops.] 

2. The Model S will be able to go up to 370 miles on a full battery charge, while the Model X can achieve 325 miles. The sturdier-looking Model Y, scheduled for Fall 2020, can go 300 miles. [Note: research any other major car company's EV claims, and all of them claim similar mileage, indicating Tesla no longer enjoys a clear competitive advantage.] 

3. The most energy-efficient cars are all Teslas. [Again, the definition of the metric is key. How does one define "most energy-efficient"?] 

4. Operating costs of electric vehicles are much less than gas cars due to the ongoing maintenance required for a "regular engine" car. (e.g., no oil changes, fewer moving parts needing replacement, etc.)
5. On self-driving capable cars: Tesla has a goal of "one million robotaxis by 2020," but still needs regulatory approval. 

6. Tesla claims to have the world's largest battery factory and discussed opening new production facilities in Shanghai, China and Europe. [From 10K, pp. 4: "We have also pioneered advanced manufacturing techniques to manufacture large volumes of battery packs with high quality at low cost.] 

7. Tesla Energy, which makes products primarily for home/consumer use, is working on integrated, renewable energy ecosystems that last thirty years. 

8. On news reports of Tesla's safety record, Musk blamed a "crazy disinformation campaign," saying every year, about 200,000 gas fires are reported in combustion engine cars [but no one seems to focus on those instances]." 

9. Tesla has discovered "two critical selling points" in a consumer's decision to buy an EV: 1) charging stations within a reasonable distance of the consumer's home; and 2) the presence of charging stations on routes drivers want to take (so they're not accidentally stranded).
10. "When you buy a car, you're buying freedom," and any unexpected repairs interfere with more widespread acceptance of Teslas (and EVs). To that end, Tesla seeks to provide insurance directly to help achieve customer satisfaction, including a mobile repair service. (Bonus: from 10K, pp. 58, "Cost of services and other revenue increased $651.3 million, or 53%, in the year ended December 31, 2018 as compared to the year ended December 31, 2017. The increase was primarily due to the increase in the cost of our new service centers, additional service personnel in existing and new service centers.") Musk cited one instance where a repair was completed in less than one hour, later disclosing Tesla needed to complete a "small acquisition" but was "close" to selling insurance. 

11. In response to a complaint that production has been battery-constrained for some time, Musk said, "I'm sometimes [overly] optimistic. I wouldn't be doing this if I wasn't optimistic." 

12. SpaceX's satellite antenna won't be linked to Tesla's cars due to the large size of the receiver required but could be used generally for under-served and poorly served reception areas. 

One last point: Tesla continues to be a beneficiary of large tax credits, some of which don't expire until 2033(!). Check out these two pages from its 2018 annual report. 
Like Musk, I hope in time, all car companies will "go electric," but I wonder if it's possible without government subsidies and pressure on local and state governments to continue to invest in public infrastructure. While Musk has prospered in California, the state government has failed to complete a high-speed rail project, leaving it behind Tokyo and other Asian cities. Low-cost insurance, if delivered effectively, will help reduce some of the burden on the poor and middle-class, but at the end of the day, if state governments are relegated to subsidizing the private sector for public needs, the future remains as it was before Tesla: uncertain, cloudy, and stratified. 

© Matthew Rafat (2019) 

Disclosure: As of the date of publication, I own one share of Tesla (TSLA) stock. 

Bonus: If you think Tesla Inc. has always been associated with Elon Musk, look up Martin Eberhard (who claims by 2020 or 2022, EVs will be cheaper than legacy vehicles) and Marc Tarpenning. Elon Musk receives credit for Tesla because his "passion" and marketing teams are able to raise money from venture capitalists more fluidly than the original founders.

Bonus: "Assuming the full electrification of the light-duty vehicle fleet by 2040, global energy-related CO2 emissions could potentially be reduced by about 5%." -- ExxonMobil's 2019 10K, page 9.