First, we kill all the lawyers

The world in general, and my industry (high tech) in particular is filled with people who think that a Better Mousetrap will change society.  Engineers are prone to the technocrat's disease: it's elegant, and insanely cool, and wildly useful to society, and so it's obvious that it will not just be built, but will dominate the market.

You'd think that one look at Microsoft® Windows® would disabuse them of that notion, but the attitude is perennial.

Via The Antiplanner, we find an example that encapsulates this fallacy to a T.



This has no hope of ever being fielded in quantity.  The reason is Lawyers.  Audi spent years defending itself against frivolous suits claiming unexpected acceleration.  It was all driver error by elderly (i.e. failing faculties) yuppies (i.e. sense of entitlement; ignore the "y" in yuppie) SWPL types.

And even so, it destroyed the Audi 5000 brand.  And this wasn't a one-off situation, a "black swan": the Fed.Gov just put Toyota through the wringer for precisely the same thing.  Of course, it was all driver error, too.

But Borepatch, I hear you say, if you could eliminate driver error, that would solve the problem, wouldn't it?

Nope.  This isn't an engineering problem.

Consider a hypothetical future, where engineers design self-driving cars.  There's little doubt that these cars will be safer than today's cars.  There's also no doubt at all that the design and implementation of the new system will have defects.  This is, after all, my chosen profession, and we can take it as a give that software have bugs, and that some of those bugs will be horrendous.

Now add lawyers into the mix.  There's typically little money to be made chasing wrongful death lawsuits caused by reckless or intoxicated drivers.  The insurance payout is in part designed to make this the case, and unless it's a business involved, the return on legal hours invested in a lawsuit simply doesn't pay.

So the current high rate of death is a given, but cannot be compared to our hypothetical future.  When a software bug causes a fatal crash - and it's not a question of "if", it's a question of "when" if millions of driver-less cars end up on the road, the legal calculus changes.  Now there are deep pockets (Toyota's), and you have an actual error due to their design or implementation failure.  A jury will want to find for the plaintiff.

Enter Game Theory.  If Toyota is smart, its Product Managers will be performing Annualized Loss Expectancy (ALE)  calculations as to expectation of legal outcome.  The profit from the expected potential additional vehicle sales is almost certainly swamped by potential losses from litigation.

The benefits of the hypothetical system accrue to individuals (people who do not die in accidents caused by drivers, and their dependents) or society (higher capacity factor for existing roads).  The losses are concentrated (people who die in accidents caused by software bugs, and the auto manufacturers after the heirs of the deceased win lawsuits).

In short, why on earth would a rational company invest in this technology?  Sure, it has real benefits.  Overwhelmingly good benefits, in fact.  But qui bono?

The only thing that would change this calculation would be for government to shield the manufacturers from judgement.  I'm not at all sure that this would be A Good Thing, as it is somewhat of a disincentive for product quality investment.  But it doesn't appear that the Fed.Gov is interested - after all, they just put Toyota through the wringer for something that wasn't Toyota's fault.  And The Antiplanner himself captures the problem with the current ruling class and their attitude towards SWPL prestige projects:

Apparently, all it takes is a totally unrealistic vision to persuade people supposedly as sophisticated as the editors of the LA Times. The truth is bullet trains are far more expensive than airlines (75 cents vs. 15 cents a passenger mile); Amtrak’s safety record is far worse than the airlines (1.4 vs. 0.1 passenger fatalities per billion passenger miles); and cleaner depends on the energy source (and powering trains with renewable energy won’t help much if all those trains do is displace some other energy consumer who therefore relies on fossil fuels). As for “reducing reliance on gas-guzzling automobiles,” the state’s own extremely optimistic numbers show that California high-speed rail won’t displace more than 2 or 3 percent of the state’s auto driving; and by the time it is built, autos won’t be guzzling that much gas anyway.

But they don't think about this, and don't want to: Don't bother me with your pesky facts, I'm building the New Progressive Jerusalem.

The ruling class is happy with their false, feel-good stories.  It's not an engineering problem, it's a philosophical world-view problem, and it won't be solved by private enterprise.