Kort Explores conundrums of the human condition

I've had some trouble with my car recently, and it continued long enough to provide a number of unsought insights into the role of expectations in perceptions and reactions to failures.

Sometimes my car would would work fine, sometimes it would stall and leave me stranded. Attempting to use the car became a very stressful activity since there was an unpredictable risk that it would stall and refuse to restart for several hours. Power from the engine was unreliable - at times I couldn't be sure whether the car would accelerate or stumble when I pressed on the accelerator. Current roads and traffic don't accommodate unreliable and/or unpredictable vehicles anywhere near as well as they used to.

Traffic around Portland is a lot less aggressive than many large cities, but it still requires some degree of reliable responsiveness to merge into traffic, get across intersections before the light changes, etc. Dangerous situations can all too easily be created if a car doesn't perform as expected - both as the driver expects, and even more importantly as the other drivers expect.

It can be surprisingly difficult to get people to respond to a danger that is outside of their previous experience and expectations. I became especially aware of this problem back when I was skydiving. People don't expect things to fall out of the sky on them, and so awareness of what is above them is very low on their priorities. This oblivion towards "up" can be a major problem for a skydiver about to land among a group of people whose awareness of their surroundings only extends as high as eye level.

I remember making a jump over Cannon Beach on the Oregon Coast. It was nearly impossible to get the people on the beach to look up and see that we were about to land on them. They obviously heard us yelling for them to "look out" and and "get out of the way", but it didn't occur to them that the sound had come from above. The idea that a bunch of skydivers were descending from the sky was so far outside of their expectations that it took a long time for them to realize what was actually happening - in a few cases too long for there to be enough time left to step out of the way.

Variations on the same theme of oblivion to the unexpected are common among drivers. Cars have become so reliable that there is a pervasive expectation that every car should be capable of keeping up with the cars around it. The assumption then becomes that any variation from the norm is due to the driver's intentional or negligent actions. For example, if a car fails to move within a reasonable time of the light changing, the automatic assumption is that the driver isn't paying attention, not that the car is suffering from a mechanical malfunction.

These days if a car in the left lane of a freeway slows, the first impulse of the surrounding drivers is to get around the driver who is unwilling to keep up with the flow. Oregon has a lot of inconsiderate drivers who go out of their ways to obstruct others. The expectation of the overtaking drivers is not that there might be an emergency and they need to get out of the way. The typical reaction of overtaking traffic is to try to flow around in the right lane, which in the case of a real emergency, blocks the slow car's access to the outer lane and the shoulder, and thereby substantially increasing the danger. If it can't get to the shoulder, the broken car will be forced to stop in the middle of the freeway causing a traffic jam. Even worst, the traffic trying to flow around the slow car may obscure the view of an overtaking vehicle until it's too late for an inattentive driver to avoid rear-ending the slow car. This can in turn cause a chain reaction collision involving dozens or even hundreds of other vehicles.

Drivers no longer expect a car to falter starting up from a stop light, or slow down climbing a hill. Being suddenly forced to drive slower than expected can easily create a more dangerous situation than moderately exceeding the speed limit because it is the more unexpected condition. It takes a great deal more white knuckled anticipation to operate an unpredictable vehicle in traffic without causing an accident or traffic jam - stress that is added on top of the "normal" stress of driving in traffic.

Quite often trying to deal with the erroneous expectations of others can be far more difficult than dealing with the underlying problem.

Our expectations seem to largely determine our psychological and emotional reactions to a failure. We are traumatized more by an unexpected failure than when the failure is to some degree expected. It isn't uncommon for most of the adverse effects we experience to be due to the failure being unexpected rather than the direct effects of the failure.

Drivers a century ago wouldn't have been anywhere near as stressed and traumatized by the performance of my car. I can remember back when the levels of reliability we take for granted today were beyond the capabilities of even the most expensive cars. My car certainly doesn't fall into the premium category. I needed a small station wagon in 1988 capable of hauling two complete computer systems in their shipping boxes. I bought my VW Fox GL largely because a brand new one was cheaper than a 10 year old used Volvo wagon.

The odometer of my 21 year old car has just recently ticked over 100,000 miles, and its drive train is still almost entirely original parts. It doesn't burn oil and continues to pass its emissions tests. It's only received minimal maintenance - mostly just regular oil changes. I've had the brakes and tires replaced a couple of times, but it still has its original clutch. Most of its problems have been the result of a lack of use, not wearing out. It isn't unusual for it to sit for a week without being driven. In spite of the benign neglect, it has (nearly) always started when I turned the key, and reliably took me where I wanted to go. This long established expectation of reliability was why it was such a traumatic and stressful shock when the car suddenly became unpredictably unreliable after two decades of reliable performance.

It wasn't that long ago when even the reluctant performance of my car as it sputtered and faltered would have been cause for elation and celebration that it ran at all. A century ago breakdowns were an expected aspect of motoring. There was a time when a car that could be driven for more than a couple of miles without breaking down was considered phenomenally reliable. In the early days of automobiles, it wasn't a good idea to drive anywhere from which you couldn't walk home when/if your car broke down. In my youth, a car with 100,000 miles was pretty much worn out, and had probably had a major overhaul along the way to have lasted that long.

While failures have become much less common, the typical user today is much less capable of fixing a problem than with previous generations of technology. Most early drivers were by necessity also pretty fair mechanics. Most things that could go wrong could be fixed along the side of the road with only a handful of tools, some bailing wire, and maybe a bit of chewing gum. Race cars had two seats so that a mechanic could ride along to keep the car running.

Back in the late 60's I had multiple cars, trucks, and/or motorcycles with the expectation that one or more of them would be in pieces in the garage at any given moment. The objective was to have enough alternatives that at least one would be usable to go get the parts needed to fix the others. The need to tow my broken cars home was sufficiently routine that I had my own tow bar. I also had ramps to load my broken motorcycles into the back of my "re-purposed" hearse to haul them back home for repairs.

These days I have neither the equipment nor the need to tow my one and only car home on a regular basis. That also means that if the need unexpectedly arose, I wouldn't be able to do it even if I wanted to. How many drivers today have more than a vague idea of how their car works, let alone the skills and knowledge needed to fix it in the event of a breakdown?

It wasn't that long ago that computer users expected to reboot their machines every couple of hours. Constantly saving your work because your computer might fail at any moment was just routine. My daily use machines now all use Linux, and run for months without crashing or needing to reboot. At the time I'm writing this, Intertel's server has been running continuously for 295 days since the last power failure that was long enough to exhaust its UPS battery. I've had a Linux computer run 430 days without being rebooted. Now that needing to reboot has become a rare event occasion, the psychological effects when it does happen have become much greater than when rebooting was an everyday expectation. I'm shocked and outraged that this could be happening to me, rather than just routinely accepting the expected inconvenience.

Joseph Stalin, one of the great monsters of human history, is reported to have observed something along the lines of: "a single death is a tragedy but a million deaths is just a statistic".

As our technology has become more reliable, we've become more aware and focused on each individual failure. In a paradox of human nature, as failures become increasingly unexpected, each one generates greater adverse reactions than multiple expected failures. The thousands of brave men who died in the early days of aviation are just a statistic in the history books, while the deaths of the seven members of the Challenger crew continue to be a national tragedy five years after the unexpected event.

Unfortunately, protecting against increasingly unlikely causes of failures can increasingly obstruct meaningful progress in the original purpose of the technology. Consider the rapid rate of progress in the early years of aviation when it was recognized as a dangerous undertaking, and crashes were an unfortunate but expected occurrence. Manned space flight has largely stalled today because we have become so risk adverse. There have been so few recent fatalities that the program has been increasingly crippled by an expectation of absolute safety.

As technology increases in complexity, it becomes more vulnerable to cascading failures that can greatly magnify the effects of a original relatively minor failure. In my case, most of my car was still capable of functioning, but the overall usability of the system was effectively eliminated by what turned out to be the failure of a single small component. There is an old saying, "for want of a bolt, the ship was lost". However, finding which small component was actually at fault ultimately required special diagnostic equipment and substantial specialized knowledge.

We take for granted that our technology will function as expected even as its complexity increases. It should be a source of pride and not a little amazement that complex modern technology works at all, let alone delivering unprecedented levels of reliability. But instead of profound appreciation of the progress we've made, users today are likely to take for granted that technologies they don't understand will function as they expect - even when that expectation has no basis in reality. Government regulations require increasingly absurd warning labels in a futile attempt to protect uninformed users from misusing products in irrational ways. Engineers are increasingly focused on protecting against increasingly unlikely unexpected events rather than contributing to the technological advancement of mankind.

In my childhood, working on cars was a primary means of learning how things worked, and arguably contributed substantially to technological progress. Today, engaging in the kinds of learning experiences that were common in my day has been increasingly prohibited by government regulations. Kids today have been deprived of a highly effective way of learning about the world around them. Who knows how many quantum leaps in engineering have been stillborn because the kids who would have invented them weren't allowed to get their hands dirty and learn what actually goes on under the hood.

Many of the technologies on which our modern world depends are becoming black boxes that few understand and even fewer can fix when they fail. These black box technologies have become more reliable, but at a cost of stifling innovation and understanding. When one of the increasing numbers of black boxes around us unexpectedly fails, we react in ways that compound the adverse effects - often causing ourselves and our society far more harm than the actual effects of the failure. Has the trade-off really been a good deal?

We've become so obsessed with magnifying the perceived impacts of unexpected failures that we are increasingly obstructing further progress. Is our growing obsession with fear of the unknown and unlikely destroying our ability to succeed? Is our fixation on absolute safety and reliability worth the cost?