Month: August 2018

How Much Ruin Is in a Nation?

If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:

Or download the MP3


It is said that after General Burgoyne’s surrender at Saratoga during the Revolutionary War that a young nobleman, named Sinclair, came to Adam Smith, and declared that the surrender would be “the ruin of the nation.” To which Adam Smith reportedly responded, “There is a great deal of ruin in a nation.” Meaning, one supposes, that nations are sturdy things, and it takes more than you probably think to kill them off, or even cause them long term inconvenience. And certainly the UK remained a great power for many, many years after this.

In our own day we have a lot of Sinclairs running around saying that this or that will be the ruin of our nation. In the past I think it’s fair to say that most of the Sinclairs came from the right side of the political spectrum. But, lately, global warming and the election of Trump, among other things, have given the Sinclairs on the left greater visibility. Despite all this panic from both the right and left, I think Smith’s statement is still essentially correct, “There is a great deal of ruin in a nation.” And I think we’re a long way away from the US collapsing in spectacular fashion, if indeed it ever does. Most nations go out with a whimper, not a bang.

There is one possible exception to this: nuclear weapons. As usual they’re a huge wild card. I don’t think some limited exchange would cause the nation to collapse (though given the extreme reaction to 9/11 who knows.) But certainly if there was all out war between the US and Russia you’d be shocked if the US wasn’t fundamentally altered in some critical aspect. (Loss of territory, change in system of government, etc.) And to be fair to the people panicking over Trump, we have foolishly given the current president absolute control over the nuclear arsenal. All of this is to say, that I’m not going to rule it out, but that’s a subject for a different post (one I’ve already written.)

On top of this, there are also of course natural disasters, the eruption of the Yellowstone super volcano, a new Carrington Event, etc. But I want to talk about what might happen in the normal course of things, in fact I want to talk about what might happen without any huge black swans. Which is not to say there won’t be black swans, there definitely will be (positive and negative) but to look at how bad things can get even without them.

I was prompted to write on this topic when I read a recent Bloomberg article by Tyler Cowen: The Decline and Fall of the American Empire, where he essentially takes the same approach. He lists numerous challenges America will soon be facing. None of these challenges are surprising, or especially dramatic, but neither are any of them simple to solve, or easy to avoid. I’ve decided to go through his list, make some of my own comments, and then, at the end, add one item of my own as well.

He starts off by assuming that we get through the Trump presidency without a major constitutional crisis. I think that’s the way to bet, but he’s already assuming a rosier outlook for the future than many people, who can’t see farther than 2020. And it’s after 2020 when things start to get difficult, because it’s only then that recent financial trends really start to reach a tipping point. Cowen’s summation of the problem is excellent:

In recent years, the underlying rate of productivity growth often has been about 1 percent, and rates of economic growth are not even half of what they used to be. Meanwhile, America will have to increase taxes or reduce spending by about $2,200 per taxpayer per year to keep the national debt-to-GDP ratio from rising ever higher, and that figure predates the Trump tax cuts. To fund that shortfall, the U.S. will cut back on infrastructure maintenance. At least one-third of this country will end up looking like — forgive the colloquial phrase — “a dump.” The racial wealth gap will not be narrowed.

This takes us back to the question of whether national debt matters. I have argued strongly that it does, and Cowen seems to think so as well. Though as he mentions, if economic growth was better this might not be a problem, and indeed GDP for the second quarter of this year was a blistering 4.1%. Which Trump took credit for of course, and maybe he should, I don’t see many other great explanations for it out there, and certainly it’s the opposite of what many economists predicted would happen after Trump was elected. But for the majority of people who don’t believe Trump is going to usher in a new era of consistently high growth, and who also believe that the debt does matter, then the problems Cowen point out appear to basically be inevitable.

What effect will this have? Well it depends on what ends up getting cut. Cowen already mentioned infrastructure maintenance, but even if we cut that to zero it wouldn’t get us very far. To make any significant dent you have to cut spending from military or entitlements. Cowen is of the opinion that entitlements will mostly survive, meaning most of the money will come from the military. I generally agree with this assessment, though I’m not sure it’s the right choice, at least from the perspective of avoiding big negative black swans. Allow me to explain.

If you cut entitlements, then you have a lot more poor people, and presumably these people suffer more than they would if they were not as poor. More alarmingly you might end up with people dying who otherwise wouldn’t have. So it is bad, potentially very bad. Though I think in the end it wouldn’t be quite as bad as people predict. And I would even predict there would be some long term positives in there as well.

If you cut the military, then suddenly rather than having the umbrella of US global hegemony keeping regional conflicts below a certain level, you end up with a re-ordering of regional spheres of influence. Here’s how Cowen imagines it playing out:

Over a period of less than five years, China will retake Taiwan and also bring much of East and Southeast Asia into a much tighter sphere of influence. Turkey and Saudi Arabia will build nuclear weapons and become dominant players in their regions. Russia will continue to nibble at the borders of neighboring states, including Latvia and Estonia, and NATO will lose its credibility, except for a few bilateral relationships, such as with the U.K. Parts of Eastern Europe will return to fascism. NAFTA will exist on paper, but it will be under perpetual renegotiation and hemispheric relations will fray.

I’m not sure where Cowen starts his five years. From today? From the minute we start cutting the military? Regardless, the China scenario he outlines is becoming increasingly likely regardless of what the US does, and a reduction in our military would only hasten it. But it’s his next claim that’s more alarming, that Turkey and Saudi Arabia will build nukes. Though, once you assume the decline of American hegemony, than the rise of regional powers would seem to naturally follow. And it’s hard to imagine how someone can be taken seriously as a regional power if they don’t have nukes, particularly if their rivals do. People always talk about India and Pakistan in the same breath and yet India is eight times as wealthy and has seven times as many people. Yet they both have nukes, which gives Pakistan an equality it could otherwise never hope to claim. On top of the two countries Cowen named, Turkey and Saudi Arabia, I would add the usual suspect of Iran, but there’s also a good possibility Japan would develop nukes.

Here, we can finally see why I’m not sure that cutting the military in preference to cutting entitlements is the right choice. Yes, cutting entitlements would almost certainly cause some suffering, but it’s hard to imagine how a nuclear war, even a limited one wouldn’t cause a lot more. Though in reality the end of American hegemony has to come eventually, I’m sympathetic to people who think it might be replaced with some kind of global government which will keep the peace. But nothing in the current geopolitical landscape leads me to think that has the remotest chance of happening. All of this is to say that there doesn’t appear to be a lot of good options.

Cowen moves on to talk about drugs, and while he thinks the opioid crisis will eventually abate, he worries a lot about drugs which can be whipped up in a “home lab”. I’m not sure I share his optimism about the opioid crisis, I agree it’s hard to imagine that it will continue at the same rate. Overdose deaths from Fentanyl increased at an annual rate of 88%(!!!) from 2013 to 2016. Were that rate to continue, the entire country would be dead by 2030. But even if everything were to freeze where it is we’re still looking at 64,000 deaths from overdosing on drugs per year, which is now nearly twice as many as die in car accidents every year. As far as “home lab” drugs, also called designer drugs, I’m not sure how big of a problem they will be. Obviously I’m inclined to caution, but my initial impression is that so far they (fortunately) haven’t lived up to the hype. A search on designer drug danger turns up a lot of articles from 2013, and the number three hit is from 2003, so I’m not sure what to make of that. At a minimum it doesn’t seem like we’ve reached a crisis point yet.

There was an article just recently in the Atlantic claiming that marijuana is a lot more addictive than people claim. And I feel like I should mention it in this section. Though fortunately it’s basically impossible to overdose on marijuana, it does seem like when you pull all of this together that there’s a real danger that we’re turning into a nation of lotus eaters.

Cowen goes on to talk about technology, but rather than talking about benefits of it, he chooses to focus on the potential downsides. Many people are optimistic about the future precisely because of technology, but Cowen points out that in the near term we’re more likely to see technology being used to create a dystopian police state, than a transhumanist paradise. He doesn’t mention China in this context, but that’s an example worth considering.

Those, like me, who were around during the Tiananmen Square protests of 1989, can probably remember thinking that the fall of the Chinese communists seemed inevitable. Recall that 1989 also saw the beginning of the end for the Soviet Union, the fall of all Eastern European communist regimes and the tearing down of the Berlin Wall. Change was definitely in the air. (Negotiations to end South African apartheid began the next year in 1990). And yet amidst all that change the Chinese Communists hung on. How they did that is a post unto itself. (One I’m almost certainly not qualified to write.) The important point is that somehow, despite all the technology that has come along since 1989, in particular the world wide web, the Chinese communists have continued to hold on.

It’s unclear how much technology has hindered the ability of the Chinese communists to hang on to control up until this point, but going forward there is a lot of technology which promises to greatly assist them in that endeavor. And here we turn back to Cowen:

Artificial intelligence and facial and gait surveillance will lead to unprecedented invasions of privacy, causing another 1 or 2 percent of Americans to decide to “live off the grid.” The impact of assassin drones will be curbed — by filling the skies with police drones. Public crimes will plummet, but public spaces in major cities will have a depressing sameness, due to the near-total absence of spontaneous behavior. Advances in recording technologies will make most conversations in public, and many in private, remarkably bland.

As you may have noticed, he’s talking about America, but you know that however bad it is here, it will be ten times as bad in China.

From there he goes on to touch on global warming, and much like myself, argues that “The very worst fears about climate change won’t come true.” Though, he argues that it will still be a significant drag on the economy, making the growth necessary to avoid cutting the military (or, less likely, entitlements) that much harder to achieve.

Cowen decides to end the article by tossing in some of his own petty gripes. Which include the fact that as more and more of current media is being driven by what is or isn’t available to stream, that fewer people are going to end up seeing classic movies. (Though one would think that once the copyright expires everyone will have old movies available for streaming, but maybe the cost of digitizing say, Battleship Potemkin, is prohibitive.) He also predicts that live performance of classical music will become less and less common. I can certainly sympathize with both of these complaints, though if I were to list my own petty annoyances I would probably come up with something like the proliferation giant ear gauging, or the vacuity of social media, or how the Hollywood regurgitation loop just keeps getting smaller and smaller.  But that’s just off the top of my head (and my first annoyance is almost certainly due to some giant gauging I saw just yesterday.)

The point of all of this, as I mentioned at the beginning, is that there are a lot of current trends which suggest ways in which the future could go poorly, without going catastrophically, and Cowen has done an admirable job of explaining many of the ways this could happen, but he did miss a few. This is entirely understandable. He made no claim to being comprehensive. But there’s one in particular that I think deserves a mention, and it’s one which might end up being as, if not more, consequential than the things he did mention.

This is the population growth trend in Africa. Most people will automatically translate this into a claim that there are “too many Africans”, which is not the sort of thing one talks about in polite society (and probably one of the reasons why Cowen didn’t mention it.) For myself, I certainly hope that how ever many people there end up being, both living in Africa and elsewhere, that they will grow up happy and well fed, and that they will all be blessed with the same opportunities and advantages as people born in the United States or Europe. But, I also hope that I’m wrong about the dangers of the national debt and that medicare and social security and Pax Americana will last forever. And, unfortunately, just because I hope for all these things doesn’t mean they’ll happen, though of course they might.

All of this is to say that Africa might have problems with overpopulation, just as all the things Cowen mentioned might happen as well. And this is certainly the way the trend is pointing. if you look at the official UN predictions for the population of Africa their most conservative estimate still has Africa reaching a population of just over 3 billion (up from 1.2 billion right now) with the most extreme estimate being that it will end up north of 5.5 billion (both estimates are for 2100). Now there has already been a lot of panic about overpopulation which didn’t end up coming to pass. And I hope that my fears will prove to be overblown again, but there are a lot of ways for things to go wrong and only one way for them to go right.

As far as ways for things to go wrong, well we’ve already gotten a first taste of one way they could go poorly with the recent European refugee crisis. Whether the Europeans should have been more welcoming is somewhat beside the point. Their welcome was what it was, which was somewhat mixed to say the least. And if that’s their reaction to the flow of refugees when the population of Africa is 1.2 billion what will it be like in 30 years when the best guess of the UN is that it will be twice that?

The reaction of Europeans and other people in the developed world to the refugees is what many people are focused on at the moment, and will probably be the thing most people are interested in for the foreseeable future, but in the long term what I’m really just trying to show is that we probably can’t assume that the “riches” of the developed world will be redistributed in such a fashion that all of the additional people will be taken care of. It might be nice if they could all come to America and Europe, find a place to live and a job, or if we could just increase our foreign aid, but that seems very unlikely. The people living in Africa are going to have to figure out a way to support themselves, specifically they’re going to have to figure out how to grow enough food to feed a lot more people, potentially three or four times as many people. I’m going to be honest, that seems like a hard problem, even if all of the energy of science and technology is brought to bear on the problem.

Accordingly, another way in which the future could go poorly just based on current trends (though catastrophic may be a better word) is for there to be widespread famine in Africa. As I said, it may not happen, everyone thought there would be widespread famine in India before the Green Revolution and there wasn’t. I hope that essentially the same thing happens with Africa.

It’d be nice if at the end of it all I could point out some simple things that we could do to solve the problems Cowen mentioned, but there aren’t any, which is of course why he’s worried about them. In fact one would be inclined to think they’re unavoidable. They may be.

I can offer only two pieces of consolation. First, there is of course the advice of Adam Smith, which I started the post with. There is a lot of ruin in a nation, or in a continent or in a civilization, and I’m sure we’re all farther away from catastrophe than people think. Though that’s part of the point, catastrophe may be far away, but a situation where things are just kind of “sucky” (if you need an example think, the 70s) might not be that far away.

Second, the technology optimists, of whom Steven Pinker is the best example (particularly his book Enlightenment Now, which I talked about in a previous post) will argue that technology has solved all of the problems we’ve had so far (this is debatable, but also not the time to get picky.) This has often happened through means we couldn’t even imagine a few years beforehand, and there’s every reason to think that technology will do so again. I certainly hope that it creates another Green Revolution and that there isn’t widespread famine in Africa or anywhere else. I’m less clear how technology will solve regional conflict in the absence of American power, or the related problem of US government spending, but despite that I hope it solves those problems as well. As I have often said, I hope I’m wrong, but I fear that I’m not. Or as Cowen says in closing his article:

So what in this description sounds so implausible? Is it that you think productivity growth will come in at 3 percent? That it will all be worth it because advances in medicine will allow us to stick around in decent form until age 135? That technological breakthroughs will extend the reach of the U.S. military further yet? That the Mars colony will be awesome?

Just how lucky are you feeling?

In answer I would say, I think our luck might be running out. Or to put it a different way. The harvest is past, the summer is ended, and we are not saved.


You may be thinking, “But there must be something I can do? Maybe I could contribute to something? Maybe something which calls attention to these problems? Maybe a blog or something like that?” Well you’re in luck. You may not have heard, but I just drew attention to it, and I accept donations!


Review: Bad Blood (Theranos)

If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:

Or download the MP3


I just finished reading Bad Blood: Secrets and Lies in a Silicon Valley Startup by John Carreyrou. Actually, I should correct that, I didn’t read it, I listened to the audiobook. Which, now that I’ve decided to do a review, is kind of unfortunate, because it makes it a lot more difficult to go back and find passages I might want to quote. For that reason, and others, the review I’m going to write will be more focused on the broad lessons of Theranos, than a recap of the story or an examination of specific events. If you’re looking for that, here are some other excellent reviews which you might want to check out.

One of the reviews I just linked to starts off by mentioning a quote from the back of the book, “No matter how bad you think the Theranos story was, you’ll learn that the reality was actually far worse.” I can say, having finished the book, that was definitely my impression as well. I first heard about Theranos in 2015 from of friend of mine who works at a clinical laboratory (ARUP, which was actually mentioned in the book.) He was very suspicious of Theranos’ claims, but it was still several months before the scandal broke. Still, his suspicions stuck with me, so when the story did break, I followed along as it happened. What I’m trying to get at, is that I already thought I had a pretty good idea of what Theranos had done. But no, the quote is right, in reality it was “far worse”.

So the book paints a disturbing picture, but to truly know how disturbed we should be we need to know whether Theranos was some kind of extreme outlier. Something entirely unique, and not similar at all to other Silicon Valley startups, or if it was broadly similar to every other startup, and it was unique just in that it happened to be the one that got caught. The fact that Theranos was more of a healthcare company than a software company, certainly contributed to the way things played out, but this shouldn’t make us complacent, there are more and more non-software startups, for example Uber and Airbnb.

At this point, I should mention that I do have some experience in the startup world. I spent over a year trying to raise money for my own startup, and I worked for three years in a startup that had raised several million dollars.

In the first case, we were 16 hours away from signing the paperwork on our initial round of funding when a key angel dropped out and the deal collapsed. This was in the middle of 2008, a couple of months before the collapse of Lehman Brothers, when the country was already in a recession, so I’ll let you guess how well things went after that.

As disappointing as it was to work like crazy on something for a year only to have it ultimately amount to nothing, the experience of working at a startup which had actually raised money was a hundred times worse. It was bad for basically all the reasons Theranos was bad, enough so that reading the book (okay technically listening) was somewhat uncomfortable. Now this is not to say that the startup I worked for was anywhere near as big as Theranos. Also, as I said, Theranos was a healthcare startup, which added a level of craziness and potential harm above and beyond the craziness of a software startup. But I would nevertheless say that the difference between the startup I worked for and Theranos was more a matter of scale than of kind. (Don’t worry, there will be examples.)

On top of this, I’ve had friends who did startups and their experiences were also broadly similar (i.e. horrible). When all of this is put together (and I admit it’s not a huge dataset) it’s hard to avoid the conclusion that Theranos was not unique. That there are lots of companies operating in a very similar fashion. That if, as Mark Zuckerberg said, startups should move fast and break stuff, that Theranos’ mistake may have just been moving slightly faster and breaking slightly more stuff (and, being in healthcare, the stuff they did break was more consequential.)

Determining an answer to this question is important, because if Theranos was unique than we can marvel at the story of their mendacity and subsequent downfall and then go back to business as usual. But if Theranos was not unique then there are probably other startups who are similarly playing fast and loose with the law, and most especially the truth, and we need to worry about what the consequences will be when it eventually comes to light. In support of the argument that it’s a broader problem, in addition to the points I’ve already made, there is a history of corporate hubris occurring in clusters. I’m sure everyone’s familiar with the hubris which lead to the 2007-2008 crash, and the hubris associated with the dotcom bubble. Going farther back you have junk bonds in 1989, and the savings and loans crisis in the early 80s.

By this am I implying that Theranos is just the first taste of an eventual startup apocalypse? Perhaps, and if you’re in a position to do so I would recommend making financial hedges against that possibility, but that’s as far as my prediction extends. Rather what I’m more interested in is how the various, what you might call, hubris failsafes, built up in response to past blow-ups and crashes, performed in the case of Theranos.

To begin with you would assume that the founders and company executives themselves act as a failsafe on things, at least to a certain degree. You might expect that they’d be concerned with whether they actually have a product that will work. You might further expect that they would be wary of lying or doing something illegal, because they don’t want to get caught in those lies and they definitely don’t want to go to jail. And as it turns out both of those things did happen with Theranos, not only were the lies eventually exposed, but Elizabeth Holmes and Sunny Balwani (Theranos President and Holmes’ boyfriend) have been charged with wire fraud, and could conceivably spend 20 years in jail. And yet this possibility seemed to have not deterred them in the slightest. This seems to be the trend recently and maybe I’m imagining that it used to be different, that leaders used to admit wrongdoing, even if they weren’t directly involved, and resign. But now everyone seems to fight in the most vicious fashion possible and to the bitter end. Certainly Trump would appear to be an excellent example of this.

Speaking of politics, I think founders and politicians suffer from a similar problem, in that the type of people who are attracted to being a startup founder or executive, are not necessarily the kind of people you want to be a startup founder or executive. It takes a level of rapacious greed and delusional confidence to believe that you’re going to be the next Steve Jobs, as Elizabeth Holmes clearly did. And, of course, Holmes’ story is the heart of the book, but Balwani’s involvement is, if anything even more interesting, and where things resonated the most with my experience.

One of the most interesting things about Balwani’s involvement, which I already mentioned, but which basically very few people knew about (not the board, not the investors, and not even many of the employees) was that he was Holmes’ boyfriend. This is in spite of a 20 year age gap (and recall Holmes was pretty young). This is creepy on its face and many people have tried to spin the narrative that Balwani was some sort of svengali figure for Holmes, but Carreyrou points out that many of the most egregious acts committed by Holmes were committed before Balwani came on the scene. In any event, a creepy May-December romance is not something I experienced during my time in startup land. But Balwani did remind me of some other things I encountered.

First, you’re going to run into a lot of people with money in the startup world: Angels, VCs, entrepreneurs with a previous successful exit, etc. And most of them are going to assume that they got that money by being skilled. Sometimes that’s the case, more often they got it by being in the right place at the right time, or as most people would say, they got it by being lucky. As you might imagine those who were lucky tend not to think of it that way. They assume their money is a direct result of their skills, skills which other people don’t possess. Balwani was apparently a great example of this and the CEO of the startup I worked for also fit this profile, and based on the stories I’ve heard from others it’s pretty widespread. I won’t go into the details of my CEO lest I get sued (again, but I’ll get to that.) But Balwani made his money by creating an unremarkable company and selling it for a bunch of money at the height of the dotcom boom. That’s luck, not skill.

For the people I’m talking about, imagining that they acquired all this wealth via skill rather than luck goes a long way towards creating the delusional confidence I mentioned earlier. As far as the greed part of it, while you might expect having a lot of money to reduce that, on the contrary it appears, rather, to inflame it. If they made 10 million in a single year, then surely, someone of their evident talent can be a billionaire, if they just put in a few more years.

As I said, you might hope the founders or the executives themselves would be one of the failsafes, but clearly we have created a process which basically selects, as founders, the people least likely to have any inherent checks on their ambition. This is not to say all founders are bad, but we should be more surprised when they’re not than when they are.

Of course, the company is more than it’s founders and executives. There is also the board of directors, who have, as their primary job, keeping an eye on the, aforementioned, executives and founders. In other words they are designed to be yet another failsafe. How did they do in the case of Theranos? Not great. The story of George Shultz, the former Secretary of State and his grandson Tyler was particularly fascinating. Tyler went to work at Theranos, saw how bad it was and tried to become a whistleblower, his initial hope being that his grandfather would back him up. But in the end, the elder Shultz essentially took Holmes’ side over his grandson’s. As an aside this does illustrate Holmes’ significant charisma, which as long as we’re talking about the board of directors, allowed her to change corporate by-laws to give the shares she held 100x the voting power of normal shares. Meaning, even if by some miracle the board had turned against her they would have had no power to do anything even then.

As you may have already heard, George Shultz was not the only former politician, or politically connected individual on Theranos’ board. Here’s how Fortunate Magazine described it:

We have former Secretary of State Henry Kissinger, former Secretary of Defense Bill Perry, former Secretary of State George Shultz, former Senators Sam Nunn and Bill Frist (who, it should be noted, is a surgeon), former Navy Admiral Gary Roughead, former Marine Corps General James Mattis, and former CEOs Dick Kovacevich of Wells Fargo and Riley Bechtel of Bechtel. There is also one former epidemiologist—William Foege, and, in addition to Holmes, one current executive, Sunny Balwani, who is Theranos’ president and CEO.

It’s quite an impressive group, isn’t it? But here’s what it’s not: an appropriate board of directors for a company that is valued at $9 billion. There are no sitting chief executives at other companies—a basic tenet of board best practices. There is but one still-licensed medical expert, Bill Frist (Foege, age 79, is retired). And while it’s probably useful to have a retired government official or two to teach and offer good leadership skills, when there are six with no medical or technology experience—with an average age, get this, of 80—one wonders just how plugged in they are to Theranos’ day-to-day activities.

Interestingly Fortune was the same magazine who put Holmes’ on their cover the year before this. And at that point they called the Theranos board of directors, “the single most accomplished board in U.S. corporate history.” And this illustrates the problem, not only was the board not equipped to act as a failsafe on Holmes’ and Theranos, but until Carreyrou broke the story, no one took the obvious step of pointing out that Holmes’ board was great if you wanted to skirt regulation and intimidate people, but awful if you were actually interested in designing an effective medical device. The article I just quoted from was only published after Carreyrou’s initial WSJ Piece. But it is only in light of that reporting that Fortune finally makes the same point I just made.

[The Theranos board] was assembled for its regulatory and governmental connections, not for its understanding of the company or its technology.

Of course now it’s arguably far too late. Interesting how quickly people decided that what was once thought of as a major strength of the company should have been regarded as an early warning sign that something was very wrong.

To be fair, there wasn’t much evidence of Holmes using the board to directly subvert regulations. (Though I suspect indirect influence played a large role.) But regulations are, of course, another failsafe, and eventually laws and regulations are what killed Theranos, the question is should they have done that sooner? A discussion of regulations will often devolve into a partisan fight over whether Republican antipathy towards regulation is really at the root of the whole problem. I don’t think that argument can be made here. Lots of very prominent Democrats were big supporters of Theranos, and this support extends beyond the board of directors. (Obama made Holmes one of his business ambassadors and Biden was visiting and praising the Theranos facilities just a few months before the scandal broke.) Meaning I don’t think it would have made any difference which party was in power, nor is there some repealed law one could point to as a smoking gun which enabled the whole disaster.

Actually if there was any single thing which empowered Theranos to continue for as long as it did, that thing was lawsuits, and more broadly the threat of lawsuits. And if there was any person to hold responsible for that it was David Boies, known for Bush v. Gore (he was on Gore’s side) and US v. Microsoft. For my money Boies is easily number three on the Theranos villain list and given that he escaped relatively unscathed (his role at Theranos is barely mentioned in his Wikipedia article.) And that of all the people involved he’s the one who really should have known better. Perhaps he deserves to be number one on that list.

As I alluded to earlier, my experience working for a startup also led to me being sued. The case has been settled and dismissed with prejudice. Also you can see I’ve tried to be careful to not mention any names, but it still makes me nervous to talk about. At the time I was sued I kind of figured that it was exceptional, that the guy paying for the lawsuit had made a bad decision based on bad information. But having read the story of Theranos, I get the feeling that lawsuits are more common than I thought in Silicon Valley.

The lawsuit side of the equation is interesting for several reasons. First it’s interesting to speculate how it interacts with the failsafe of government regulation. Would having more laws just give Boies more laws to sue people under? Or would it have made it easier for the government to crack down on Theranos. I strongly suspect the former, particularly given that once the appropriate agencies were made aware of what was happening, they had no problem shutting them down via the laws that were already on the books. But getting to the point where they were aware of the problems took a lot longer than it should have because of Theranos lawyers. Second, as far as I can tell, lawsuits have way more to do with how much money you have than whether you’re right or wrong. This is not to say that the side with the most money always wins, but to even play the game costs tens of thousands of dollars (in my case) and hundreds of thousands of dollars (in Tyler Shultz’s case).

Finally lawsuits, as I already alluded to, act as the opposite of a failsafe. In that they make all the other failsafes more difficult to trigger. As it turns out, despite however unethical the people at the top of Theranos were, Carreyrou tells the story of dozens of employees who wanted to make the world aware of what was happening, but were kept from doing so by the fear of a potential lawsuit. The story of Tyler Shultz facing down David Boies was one of the most gripping parts of the whole book, and if he hadn’t been the grandson of a member of the board and had parents who were willing to spend $400k to keep them at bay, it’s unlikely that he would have persisted either. Also once Boies turned on Tyler, even $400k was not enough to allow him to continue to talking to Carreyrou. He had to immediately shut up AND pay $400k.

This takes us to the failsafe which finally brought Theranos down, the press. One would imagine that once the Wall Street Journal got ahold of the story that it was all over. But it’s surprising how much effort Holmes, Balwani and Boies put into trying to quash the story even then. Carreyrou was constantly hounded by attorneys (and followed everywhere he went by private investigators). More interestingly, Holmes, aware of the story, managed to convince Rupert Murdoch (the owner of the WSJ) to invest $125 million in Theranos, and then proceeded to use that to try to leverage him into killing the story. I know lots of people don’t like Murdoch, but to his credit he apparently not only refused to kill the story, but refused to involve himself at all, and eventually lost the entirety of his investment (a drop in the bucket for him, but still.)

The final failsafe is supposed to be the most reliable of all, money, and by extension self-interest. How did Theranos raise $700 million dollars, and secure lucrative contracts with both Walgreens and Safeway despite not having anything close to the technology they claimed? Carreyrou’s answer was FOMO, or fear of missing out, which of everything I’ve mentioned so far may be the most modern phenomenon of all. And also the one I’m the least sure how to solve. It’s clear that FOMO is a major driver of bubbles, as everyone sees the upward spike and rushes in, never pausing to think if the growth is sustainable or if they’re just the greater fools. But what should be done about it? It’s a big problem, but in the very specific case of startups, I think it needs to be much easier to bet against them. And one of the things that has made that harder is the decline of the IPO.

All of this is a fairly large topic, and I’m out of space, but I will say that if I was one of the investors in Theranos, or one of it’s board members, or Walgreens or Safeway, that I would be taking a long hard look at how I ended up being so thoroughly duped and how to prevent it in the future.

As I said in the beginning, it’s not clear if Theranos is an extreme outlier, but I suspect it’s not, and I further suspect that if the many lessons of Theranos are not learned, we’re in for a lot more of this sort of thing, and I predict some of it will make Theranos seem quaint by comparison.


Much like Theranos, there are many areas where I’m incompetent. In their case it was worth $700 million. I’m not expecting my incompetence to be worth that much, but if you think it’s worth anything consider donating.


Pornography and the Time Horizon of Uncertainty

If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:

Or download the MP3


As technology advances, new innovations are periodically introduced. Sometimes these innovations are so obviously beneficial (think the smallpox vaccine) that there is very little objection to them. On other occasions, the innovation might encounter significant resistance. A group of people may assemble who feel that this new innovation is harmful, or potentially harmful. And, if the influence of this group is large enough, this resistance might get classified as a moral panic.

Past examples of this include birth control pills, television and even heavy metal music. (You may not think that technology had anything to do with panic over music, but I think you’ll find it had more to do with it than most people think.) And of course there are plenty of panics happening right now, as just one example many people are intensely concerned about the potential negative effects of social media.

Reading this introduction one might almost assume that I’m preparing to dismiss the current panics by pointing out that past panics didn’t amount to anything and thus current panics probably won’t either. And certainly there is something to that, but I think if you look deeper at past panics, and current panics, you’ll find that the people who panic aren’t so misguided after all. What do I mean by this? Well, I would argue that many, if not most, of the things people feared did come to pass, it just took a lot longer than they predicted, their very worst fears did turn out to be unfounded, and, most importantly, by the time they did come to pass, very few people cared anymore.  

To use an analogy, you might say that the kick was accurate, but that the goalposts were moved. And true, a miss is a miss, but most people end up, incorrectly, blaming the kicker without mentioning that the goal posts were moved. Which is to say that people assume all of these panics were misguided. That over and over the forces of reactionary intolerance are always being proved wrong. And that anyone who worries about something like, social media increasing teen suicide risk, is just a neo-luddite puritan who blindly opposes all progress. Rather than someone who will turn out to be, on the whole, correct, just not quite as dramatically as they would have hoped and long after anyone cares enough to do something about it

At this point examples would probably be beneficial. Of course the problem is that, as I said, while I believe a lot of the predictions have come to pass, they rarely come to pass in a dramatic, undeniable fashion. Also, as I mentioned, to the extent that they have come to pass, people may no longer view it as a bad thing, and in some cases, people will even think it’s a good thing. For all of these reasons, I confidently predict that people are going to quibble with the examples I provide, which is fine. But I think if you start by setting aside your modern prejudices and really try to get into the head of historical people and their moral panics. You’ll realize that most of them, upon seeing the world of today, would not only feel entirely vindicated, but probably, that they had not panicked enough.

To start with, let’s look at one of the items I mentioned above: birth control pills. At the time of this panic, people confidently predicted that widespread availability of birth control pills would lead to more extra marital sex. Is there any doubt that this is exactly what happened? Now you can argue that it was not as bad as they thought, or that it’s actually a good thing, or that this came to pass, but some of the other things didn’t, but this is exactly the goalpost moving I was talking about.

There have been countless government programs passed on the idea that, whatever the costs associated, they would not end up being excessive. And since at least the days of Ross Perot, people have predicted that the costs would end up being much greater than the politicians said, that the debt would just keep growing, with an associated panic about what happens if it gets too big. Now we can argue (as we have in this space) about how harmful the deficit/debt is, but is there any doubt that it is WAY bigger than any politician could have imagined even a few decades ago? As one example of this process, Social Security was never envisioned to be a pay as you go system. From an article in the Washington Post titled, Would Roosevelt recognize today’s Social Security?

When Roosevelt proposed Social Security in 1935, he envisioned a contributory pension plan. Workers’ payroll taxes (“contributions”) would be saved and used to pay their retirement benefits. Initially, before workers had time to pay into the system, there would be temporary subsidies. But Roosevelt rejected Social Security as a “pay-as-you-go” system that channeled the taxes of today’s workers to pay today’s retirees. That, he believed, would saddle future generations with huge debts — or higher taxes — as the number of retirees expanded.

Discovering that the original draft wasn’t a contributory pension, Roosevelt ordered it rewritten and complained to Frances Perkins, his labor secretary: “This is the same old dole under another name. It is almost dishonest to build up an accumulated deficit for the Congress . . . to meet.”

A great example of a prediction which came to pass, but by the time it did the goal posts had been moved.

We’ve covered examples of morality and bureaucracy. This next example combines them. People have long argued that you “get what you pay for”. Leading many people, stretching all the way back to the New Deal, to predict that welfare might act as an incentive for various behaviors we don’t want to encourage: joblessness, a decline in marriage, etc. and disincentivize things like being resourceful and personal ambition. I understand there’s an argument to be had about how much welfare contributes to the current levels of all of those things, or if welfare and these trends act entirely independently. But imagine showing someone who opposed the War on Poverty the current figures on the number of poor, welfare spending and the increase in single mothers. Can you imagine this person experiencing any emotion other than vindication?  Particularly when you consider our society’s relative affluence.

I could go on, but at this point you’re either nodding your head in agreement, or you’ve already dismissed me as a hopeless reactionary, attempting to disingenuously fight battles which were already long ago decided against me.

For those of you nodding along, (and those who are inexplicably still reading despite dismissing me) all of this is leading to a discussion of a current panic, one which is being dismissed in the same fashion as all the previous panics.  But one where I think, eventually, the predictions of harm being made will once again, turn out to be true.

The battle in question is over pornography, and I was inspired to revisit this issue by a recent article in Slate, Why Are We Still So Worried About Wat­­ching Porn? subtitled: “Decades of fearmongering almost got porn addiction added to the International Classification of Diseases. Thankfully, the World Health Organization got it right.”

To begin with I should point out that they’re not dismissing all worry. But they seem to think that harcore porn videos should engender about as much worry as watching TV.

Even if sex-film viewing has been grossly exaggerated as a national problem, might it still be a problem for some people? Of course, just as there are excellent interventions to help reduce television viewing without invoking mental illness labels, you might want to reduce your sex-film viewing in favor of other activities you value more.

I read that as, sure it might be nice if people didn’t watch five hours of porn every day in the same way that it might be nice if people didn’t watch five hours of TV every day, but there’s nothing which makes pornography worse than binging Stranger Things. I would disagree.

As you can tell from the subtitle the article was prompted by a recent decision by the WHO to not add porn addiction to the list of official diseases. Right off the bat the article talks about, “…the shock when journalists learn that ‘pornography addiction’ is actually not recognized by any national or international diagnostic manual.” My immediate question is why are they shocked? Surely no journalist would be shocked to be told that there’s no entry in the DSM for being addicted to broccoli, and they probably wouldn’t be shocked to find out there’s no entry for TV addiction, and yet according to the authors of the article it would be basically the same thing, and yet one is obviously shocking and the other isn’t.

I would argue that they’re shocked because they know people who consider themselves addicted to pornography, and have heard stories about behavior that certainly seems like addiction. That they are shocked because it does not match what they’re seeing with their own two eyes and hearing with their own two ears. This is why it makes sense that they’re shocked, and in fact within the article they mention that a non-trivial percent of people consider themselves addicted.

Amazingly, the first nationally representative peer-reviewed study on sex-film viewing was only just published in 2017 in Australia. This study found that 84 percent of men and 54 percent of women had ever viewed sexual material. Overall, 3.69 percent of men (144 of 3,923) and 0.65 percent of women (28 of 4,218) in the study believed that they were “addicted” to pornography, and only half of this group reported that using pornography had any negative impact on their lives.

Several things jump out from this quote. First, 3.69% is not 0%, and certainly as a percentage of males it has to be higher than some of the other things which did make it into something like the WHO disease guidelines or the DSM. On top of that I think the numbers from the study are low. You’re saying, in this day and age, that 16% of men and nearly half of all women have never viewed sexual material? That doesn’t pass the smell test. And other numbers disagree with the Australian study. This survey (admitted conducted by a Christian group) claims that 13% of men admit to being addicted to porn, with another 5% are unsure (they did interview both religious and non-religious individuals). Add those numbers together and you’re in the ballpark with the number of men who smoke (16.7%).

Finally and most interestingly, they mention that this is the “first nationally representative peer-reviewed study on sex-film viewing” and that it was only published in 2017. I’m not sure if you’ve heard of the replication crisis, and the other problems with social science studies of exactly this kind, but I’m loath to take a single study and carve it into stone, as they appear to be doing. Also the self-reported aspect of this study raises a lot of red flags as well. We have one self-reported study with a male addiction rate of between 13-18% and one with a male addiction rate of 3.69%. Yes, you can certainly point out the bias of the first, but how sure are you that the second is bias free? Also, how likely is someone to self-report that they’re addicted to porn and that it’s having a negative impact? It would make sense for all the bias to be towards under-reporting on those questions.

Why would this be? Because people are ashamed of course. Why are they ashamed? Well according to the article:

Speaking to the heart of the issue, one of the biggest problems for some porn users is shame. Shame about viewing sex films is heaped on the public by the sex-addiction treatment industry (for profit), by the media (for clickbait), and by religious groups (to regulate sexuality). Unfortunately, whether you believe porn viewing is appropriate or not, stigmatizing sex-film viewing may be contributing to the problem. In fact, an increasing number of studies show that many people who identify as “porn addicted” do not actually view sex films more than other people. They simply feel more shame about their behaviors, which is associated with growing up in a religious or sexually restrictive society. Further, labeling behaviors as a mental disorder is often stigmatizing and harmful in itself, so it should only be done when strongly warranted by evidence. If we really wanted to help such people, one of the most direct ways to do it would be to normalize and validate their sexual desires, including their interest in sex films.

They appear to be suggesting that shame would be nearly non-existent without unscrupulous external organizations. Who create shame for their own selfish reasons. From this they go on to say that the elimination of shame is the best way to deal with the issue, or as they say it, we should, “normalize and validate their sexual desires, including their interest in sex films.” But if this is really what we should be doing, why is shame so deeply embedded on this issue? Even if we believe the authors and grant that there are bad actors out there using shame for their own ends (The vast sex-addiction treatment/religious-industrial complex Eisenhower warned us about!) why is shame so easy to invoke? Is it all cultural? A left over from the Victorian era, which, somehow, 50 years into the sexual revolution we still haven’t gotten rid of? Or is it more deeply embedded? Perhaps genetic? And if so, could it be that it’s been built in for a reason?

I would argue that it is genetic or common to all humans through some other mechanism. (For example, how many societies can you name where individuals have sex in public without any attached shame?) Maybe it is cultural and not genetic, just particularly tenacious culture. But in any event, from this, it follows, shame is probably there for a reason.

I have talked in the past about pornography being a supernormal stimulus. To refresh your memory, supernormal stimuli are phenomena where a species’ evolutionary preferences are unbounded in a certain direction, allowing technology to create a situation where this preference becomes counterproductive. The classic example is a species of bird which prefers large eggs, and scientists were able to get the bird to incubate artificial eggs almost as big as the bird itself, eggs which couldn’t possibly be natural. And most alarming, to prefer these eggs to the eggs they’d actually laid, eggs which might actually produce chicks.

The big question is, why would the bird do something so counterproductive? Well the bird could never lay an egg as big as she was, so there was no reason for evolution to cap this preference. (Brood parasites are obviously a possible exception to this, but outside the bounds of this post.) Turning this to humans, it obviously increases our evolutionary fitness if we have a strong urge to have sex. One aspect of this appears to take the form of an urge to see people naked or engaged in sex acts. In the millions of years humans and proto-humans wandered the African savanna there was a limit on how much of this someone might have access to. This is no longer the case. We have created giant eggs, and it’s not inconceivable that people will incubate these eggs in preference to incubating eggs with actual offspring.

As I said the big question is why do the birds do this? Obviously if this became common enough you would expect them to eventually adapt. Through language, memes and culture humans have the means to adapt to things more quickly than other species. Is it possible that shame is humanity’s adaptation to this supernormal stimuli? Recall that things like prostitutes and similar erotic “technology” has been around for a very long time.

To be clear about all this, I’m not saying that pornography is definitely a supernormal stimuli. I am saying I think that’s the way to bet.

I’m not saying that porn addiction is definitely a thing, and equivalent to smoking in its harmful effects, and something which should be included in the DSM and the WHO’s lists of diseases. I am saying that I am very doubtful that pornography isn’t a problem for a lot of people, even after you eliminate the supposed effects of externally generated shame.

I’m not saying that the authors of the article aren’t trying to be genuinely helpful. I am saying that they overlook several very large factors, which should make them far more cautious in jumping to the conclusion that “pornography” is essentially harmless.

I’ve already mentioned several factors they’re overlooking. First there’s the issue of the shifting goalposts. People say something will happen, it does, but it takes so long that no one cares anymore. We touched on how important this goalpost shifting is, which is a big topic, but if nothing else, in the here and now, we should give more weight to the predictions of those who are alarmed.

I also mentioned how little data there is. The article admits that one of the first studies done (certainly the first study they find credible, there was the other survey I mentioned) was done in 2017. Viewed in light of things like the replication crisis and the other problems currently affecting social science I don’t think their data is as ironclad as they want you to believe.

This all leads into another issue I think they’re overlooking. Not only was the first study done in 2017, but the entire phenomenon is incredibly recent. They specifically mention “sex films”. Well as I argued in another post, the ubiquity of “sex films” is basically 10 years old. Are they really saying that they can accurately predict the long term consequences of a technology that’s been around for less time than Gray’s Anatomy and NCIS?  And speaking of TV, what would predictions made after it’s first 10 years of existence look like? How closely would those predictions match what TV actually became?  The answer is, “not very close at all.”

I think in the case of pornography our predictive power is going to be even worse, when you consider things like VR pornography and sex robots (which came up during the recent discussion of incels.) And to be clear, I personally have a hard time wrapping my head around the idea of sex robots becoming very common, but it is a wildcard, and if the authors get their way and manage to banish all shame, maybe that’s all that stands in the way of their widespread adoption.

When you consider all of these things, it takes an exceptional amount of hubris to dismiss all of the concerns about pornography as groundless. And all of the associated shame around the topic as counterproductive and existing only because of external actors with harmful motivations.

In the final analysis it comes back to a consideration of costs. What does the situation look like on the balance? Does the pornography status quo, with the ubiquity of porn on one side, and the shame on the other side cause more net harm or more net benefit? The article appears to be making the argument that there’s no downside to the ubiquitous porn and only the current and potential harm that comes from the associated shame. First, as I argued, I don’t think we should be so quick to assume that shame is a useless, entirely negative emotion, with no long term benefits. And I’m inclined to believe further that even if there is some short term limited harm, that in the long run there could be big benefits.

Maybe there are some people who react negatively to shame and instead of improving their life they just sink into a deeper funk, but my guess is the number of people who don’t do a bad thing because of something like shame vastly exceeds the number of people who just get worse.

In the end what I urge most is caution. I understand that eliminating pornography from the internet is beyond quixotic, but I think that declaring all activities related to “sex films” to be “normal and healthy” (as the article does) after a scant handful of years is similarly insane.


Just like all those people in the past, eventually I’m going to be right. And yes, by that point people might not care, and also, I may no longer be around. So if you suspect that I’m right, you’re only choice is to donate now.