The Futility of Problem Solving

I happened upon an article on Unz titled The Idiocy of the Average - and Why it Matters. I am unfamiliar with the author, but a scan of his history shows he is prolific over a number of interesting topics. I'm going to add him to the sidebar.

The post is sobering, quantifying the large numbers globally that are unable to perform basic data comprehension and analysis. While I recommend the article in its entirety, here's an example of the hardest of the six levels of questions asked in surveys by the Programme for International Student Assessment (PISA).

The test was administered to 15-year-old students across the globe. (Notably missing is Africa, where only Tunisia - whose population is of Arab and European descent - is listed. No sub-Saharan countries are represented by the survey.) Results show that, in all countries outside east Asia, the number of students proficient in level 6 questions is less than 10%, and mostly less than 5%. The USA only measures at 2%! I find this statistic hard to believe. Perhaps the organization is in the business of producing alarming results for its own benefit. Still, the Asians do significantly better.

The example question shown above amounts to making some observations and computing the rate of something. It's a skill that we'd expect a good mechanic to possess. That makes it an interesting benchmark for comparison. IQ and national economic success are well correlated. In fact, your income level is six times more dependent on the national average IQ than your own IQ. The article mentions that the national metrics are even more sensitive to the "smart fraction" than the overall average. As a nation becomes more technologically advanced, the extra impact of the smart fraction should diminish because the tasks of the technician become more complex, and more technicians are needed.

There's a belief among techno-utopiasts that technology will solve all our problems. It's hard to see how technology could solve the problem of widespread technological illiteracy. Illiteracy isn't even the right word, it's more like...incapability. If someone can't conquer the rate x times = distance equation, where the numbers have been pre-selected to be trivially calculated, it's not clear how they would have the cognitive wherewithal to understand a rear differential on an F150, to manage logistics for a medium-sized company, or analyze market trends and financial figures. In short, to do the things that the modern economy requires.

Certainly, much technology exists to make the job of the technician easier. Diagnostic machines analyze vehicle issues, for example. These things have a tendency to make the easy problems easier, and the hard problems harder, because, when the diagnostic tools fail, the mechanic must troubleshoot a system that is very complex, and which he does not dive into often enough to be well-acquainted with. We see the same in software. Tooling improves efficiency, but can cause real headaches when it fails. So a shop might get by hiring a number of wrench (or code) monkeys, but it is essential that they have at least some competent employees to handle the inevitable difficult issues that arise.

Some people believe that eventually even white-collar jobs, like software development, will be replaced with artificial intelligence. The longer I work in the field, the more I believe that to be a pipe dream. Mostly what programmers do is to translate business needs into a technical solution. How would the robots be given the business requirements? As best I can tell, the only thorough description of a firm's business logic is the software itself. By the time we've described to the machine what the business should do, we've already written the code! Auto mechanics won't go away either. Robots might be developed to do some handy tasks, like removing rusty bolts from tight spaces, or even to entirely remove the starter from a 97 Dodge Dakota. But they'll only ever be a fancy tool to aid the mechanic, and will require their own programming and maintenance.

When I was in grad school, one of the PhD candidates was working on applying artificial intelligence to programming. He quickly realized that writing code was far too complex a task, so he reduced the scope to modifying code in response to runtime exceptions, such as correcting off-by-one errors, which frequently occur because most programming languages count starting at zero. The efforts bore more frustration than success. Even simple code changes could not be reliably automated.

We aren't going to automate ourselves out of our own stupidity. Which is good, of course, or there would be no need for us at all. Still it raises problems of only slightly less significance than the eradication of all humans by our robotic superiors. Conservatives imagine that anyone can lift themselves up by their bootstraps to prosperity. But people have only so much potential. The more advanced we become, the more people that get left behind entirely. The military has an ASVAB cutoff that equates to an IQ of about 85. Below that, they reckon, and a person's ability to contribute is not worth the cost. What do we do with such people? Feed them? Feed them to the pigs? Similarly, the liberals think anyone can be educated to prosperity. Which is funny, given all their efforts to destroy education. Apparently, only 2% of Americans have strong mathematical abilities after 10 years of STEM-obsessed government education. To the left's credit, they've noticed the disparity, as evidenced by their constantly blaming it on oppression by whites, men, Christians, etc. The conservatives are much less wrong than the liberals on the matter, but we are still left with a big problem. And, as we know, We Must Solve Our Problems or it's Just Going to be Communism.

On the other hand, maybe that post - written here over a year ago - isn't accurate. It's starting to look like it'll be communism whether we solve our problems or not. Look at the stats for level 3 problems. The provided example can be solved by multiplying three numbers. 52% percent of American students fail the test! Yet they all will get to vote. The results speak for themselves. Some of these Congressmen are so stupid it hurts. Maxine Waters would probably pass the level 1 test and proceed no further. That would mean that 26% percent of American voters can't even tell she's an idiot (and a much higher percentage, it would seem, in her California district). Does it really matter that much if we solve our problems? Look at Trump. A record-breaking economy that the Democrats said would be impossible, taxes cut, ISIS defeated, North Korea denuclearizing, NAFTA renegotiated to the benefit of both the US and Mexico, NATO countries starting to pay their fair share for defense, Congress working to pass their first true budget in two decades. The man slays problems for breakfast. And yet, it's not clear now whether Republicans will even hold the house in November.

A democratic government that grants a universal vote to all citizens (and non citizens), where the average voter can't solve even very simple problems, will eventually be brought under by its own inability to solve problems. It's inescapable. There is no technological solution to this problem.