Different people’s brains seem to me to be predisposed to pay attention to different kinds of stimulation. My musically inclined wife, for instance, is much more attuned to sounds of all types than I am. This makes it easier for her to identify music from just a few bars, but also makes her much more annoyed by stray sounds that I typically ignore. For example, when commercials come on the TV, she is very keen to “mute” the TV whereas I simply mute them in my mind (or at least I think I do). So, when she told me that I “had to” fix our doorbell right away, at first I had no idea what she was talking about.
“That beep!” she insisted. “Can’t you hear it? That doorbell is driving me crazy!”
After calling my attention to it, I also heard the beep. The doorbell was not something that we had installed. It came with our condo and up till now had been working just fine. Now, it appeared to be hell-bent on incessantly going “Beep! Beep!” Admittedly, it was annoying. Not so annoying as a failing smoke alarm. At least this was going off in the middle of the day whereas failing smoke alarms are not only much louder but scientifically designed to go off at around 3-4 am in the morning. I suppose on rare occasions, they do go off at other times, but I’ve never experienced that personally. Best of all, smoke alarms have directions printed right on the alarm in tiny white on white font. Seriously? You couldn’t afford to pay for .0001 cents of paint to make it legible? But enough of badly designed smoke alarms.
Let’s return to my wife’s request to fix our doorbell. I got out the toolbox and easily removed the screws over the housing. Inside were minor electronics connected with three wires to the house electricity. There did not seem to be a dying battery at fault. I had no idea, and could not decipher which wire would turn off the alarm. So, careful to touch only the insulated rubber guards on the wire snippers, I cut one of the wires. In response, I heard, “BEEP! BEEP!” Well, that didn’t do the trick. I cut another wire. “BEEP! BEEP!” Damn. Okay. I will have to cut the third wire. No battery. No electrical current from the house. Goodbye annoying beep. I cut the third wire. “BEEP! BEEP!”
What? Unlike my Dad, I was never trained as an electrical engineer, but I do know that a completely open circuit without power can’t keep “working.” At least not for long. A capacitor can hold a charge. In old time TV’s you had to be very careful. You couldn’t simply unplug the TV and start working on it right away. The large TV “picture tube” for instance, held a considerable charge until you grounded it against the chassis with a screwdriver. But there’s no way the doorbell could still be making noise.
Eventually, we discovered that there was nothing at all wrong with our doorbell. Well, to be more accurate, there had been nothing wrong until I cut every single wire. The noise source was something else entirely. Years earlier, we had attended a Dave Pelz golf academy focused on “the short game” and had been given a very cheap electronic metronome to help us learn a smooth rhythm on the putting stroke. We hadn’t ever used it for that purpose and had forgotten we even owned it.
But that’s what our lovely, lively cats are for! The cats had managed to turn on the metronome and then carefully and meticulously slide it down into the small slice of space between our piano sounding board and the wall that separated the kitchen from the dining room. Voila! A nice loud “BEEP! BEEP!” sound.
Looking back on the incident, I can’t quite reconstruct why we thought this was a doorbell. It didn’t actually sound like our doorbell. Well, nothing actually sounded like our doorbell because now it didn’t make any sound at all. I had cut all the wires that would enable it to work. But it didn’t even sound like our doorbell used to sound. Somehow, we had gotten sucked into a particular framing and formation of the problem. That specific way of approaching the problem led us down a “garden path” that not only had no possible chance of solving the real problem; it also had negative (and unnecessary) side-effects such as ruining our doorbell. Sadly, even two supposedly “well-educated” people found it all too easy to go down that “garden path.” This brings me to “Horizon University.”
Articles that claim to calculate the “best” University for you to attend have grown up like ragweed in the last few years. What irks me about such articles is not that they rank order university programs according to the average “Return on Investment” of graduates, but that they don’t even seem to acknowledge that this is only one of many criteria by which such programs could be ranked. They too, have gone down a very particular garden path when it comes to defining the “goodness” of education.
Instead of an undergraduate program that is essentially a high level trade school aimed exclusively at getting you the highest paying job, let’s imagine a University with a different focus.
Consider a University where students focus on seeing things in different time perspectives.
Maybe it doesn’t need to be an entire university; perhaps a department or a course. But somewhere along the line, it seems absolutely critical to me that people receive more training in taking a flexible view, a broad view, a long or short view, a loving view, a defensive view. In my experience, people often have one particular way of approaching a particular type of problem. In extreme cases, people approach every problem the same way. Sometimes that one way works extremely well. More often, it works pretty well. Sometimes, it is more of a hindrance than a help. And then, every once in awhile, it results in an unmitigated disaster. And, that’s true for everyone on the planet so long as you stick to one approach for every single problem.
At “Horizon University” you would not take a calculus class or a psychology class or a creative writing class. Why? Because it is all too tempting — indeed probably necessary in order to pass any such course — to use your knowledge of that particular course, using the methods of that particular course. You do not answer a calculus question with an insightful essay on the probable family dynamics of Pascal’s family; not if you want to pass.
In real life, a particular problem might require only calculus, or only creative writing or only psychology. More likely it will require some combination of these and many other skills. It will most likely be solved, not by you alone, but by you in combination with a team diverse in almost every dimension imaginable.
At Horizon University, people would be guided in every aspect of problem solving which includes the extremely important and seldom taught skills of problem finding and problem formulation. These are the hardest parts; they are the least taught parts; indeed, they are the least understood parts of the overall problem solving process.
Let’s take an example puzzle: “There are 435 people in the US House of Representatives. What is the probability that at least two Representatives share a birthday?” I have given this problem to a number of people. After a few moments thought, most smart 10 year olds can solve it. Adults have more trouble. Adults who have taken a college course in statistics however, typically have the most trouble of all. When such an adult hears this problem, they are immediately reminded of the so-called “Birthday Problem.” Counter-intuitively, it turns out that even a small group of 30 people is more likely to have at least one shared birthday than not. A ten year old is unlikely to have heard of this problem, so they think about the 435 people in the House of Representatives for awhile and come up with the correct answer. A statistics-trained adult however, is likely to say something along the following lines, repeated more or less verbatim from someone attending at a party organized by my office mate at the University of Michigan.
“Ahem! Well, this is the famous ‘Birthday Problem’ and, having just received my Ph.D. in statistics, it would be fairly trivial for me to answer this if only I had access to some logarithm tables. (This was long before hand-held internet access). I had happened to notice that my office mate had log tables so I escorted this guy to them and said, “There you go! Knock yourself out!” I went off to enjoy the party while he spent the next few hours muttering in a corner trying to make good on his boast. I checked up on him later, but he still insisted he had almost solved it.
His insistence that he knew enough to solve the problem and his persistence in tackling it with the same method over and over is one of the things that scares me about the coming ubiquity in “Artificial Intelligence” especially as it intersects with the “Internet of Things”, “Driverless Cars”, and “Intelligent Agents.” It isn’t so much that people won’t make perfect AI systems for a long time. It’s that people will make imperfect AI systems and insist that they are perfect. In other words, hubris is one of the human failings that can be greatly amplified by Artificial Intelligence.
We see this kind of hubris is all sorts of software systems; indeed, it isn’t even limited to software systems although the absurdly short development cycles of software tend to make it more evident there. For example, Microsoft’s Windows 7 had over 2000 bugs.
Bugs, of course, are not limited to Microsoft products. Here’s a list of recent bugs in the MAC OS.
“Bug” is a general term, of course, and there are many varieties. One of the “minor” kinds of bugs are usability bugs. For instance, I recently signed up for an alumni site. They asked users like me to enter the name of the University of my advanced degree. Instead of allowing me to type in the University, however, I had to use a pull-down list. This alphabetical list had over 2000 entries. But where is “The University of Michigan” to be found in an alphabetical list? Looking at the names of other universities showed no consistency whatever. It might be under “T” for “The University of Michigan.” It might be under “U” for “University of Michigan” which might be abbreviated as “U” or “Univ.” and it might be listed under “M” for “Michigan.” It wasn’t under any of these. So far as I could tell, The University of Michigan, one of the top-ranked universities in America with a current enrollment over 44 thousand wasn’t listed at all. You could call the omission of this particular university a “bug” but the more fundamental bug is why they are using a pull-down list to have users select among thousands of items. No-one thought through the fact that new universities arise; they merge; they fold. In addition, there is no obvious single way for them to be listed. But all of these errors in design thinking pale in comparison to the one that prevents the user from simply typing in the name of their university. Not only have the designers and coders of this software omitted an important option; not only have they chosen an inefficient way to enter the data; beyond that, they are so cock-sure of themselves that they have not even provided an alternative input method.
You might argue that subsequent data analysis will be easier if everyone chooses from among a fixed and finite list than it would be if people could type in whatever they wanted. True, but if that’s really the argument, then you are saying that your time and convenience are more important than those of your users. That’s too gigantic an error to be labeled a “bug.” It’s much more fundamental.
If you think I’m exaggerating the scope of software bugs, you might want to check on the Wikipedia entry of known and severe bugs in a number of different fields of human endeavor.
If Horizon University does a good job, its graduates will likely produce fewer bugs, but more importantly, they will be willing to admit the possibility that their code is buggy. Of course, bad design is not limited to software. Shelves of every store abound with poorly packaged items encased in nearly impenetrable plastic. Many roads are equipped with road signs that cannot be read at night. Processes are designed without feedback on whether they work. The crucial point here is not that humans make mistakes; obviously, they do. The problem is thinking that because you’ve learned a particular method or way of thinking that method is also capable of solving all problems; that your way of thinking is the only way there is.
Let’s return to the poor guy who spent the entire party at the University of Michigan pouring over my office mate’s log tables. He was not so much unable to apply the methods he had learned; it is just that the methods he was attempting to apply were not applicable in this case. There are only 365 days in a year (or 366 if you count leap years). But there are 435 people in the House of Representatives. So, even if the first 366 people you looked at happened to have different birthdays, the 367th would have to match someone.
At Horizon University, students would be taught a variety of methods for each part of the problem solving process. These methods would not be taught in a series of lectures. Rather, from the beginning, students would begin working on individual and group projects of their own defining. They would have access to a variety of experts including many generalists on site as well as remote experts available at varying time scales. They would hear from and see in action a wide variety of ways of attacking each problem. They would learn to respect other ways of looking at problems, not just the one or few that they themselves chose to focus on.
Everything in life is not about solving problems however. It is also important to discover and learn about the things that give you the most joy. For some people all of those things will be closely related to problem solving. But for others, many or even all of those joy-inducing activities will not really be about problem solving. They may want to hone their skills in writing, painting, music, choreography, and so on. Perhaps they will earn enough money to get by without another job and maybe they won’t. A few will find a way to use those skills as part of a collaborative problem solving endeavor. Others may find teaching their skills to others is a good way to keep their own skills sharp for their creative work.
At Horizon University, various activities and architectural features would encourage people to communicate and interact with people across the entire variety of interests. In the short term, this would be beneficial to the individual because all their project work would require a broad range of talents. Of course, in the longer term, the benefit would be understanding the value of all kinds of knowledge and skill rather than just the one that they happened to choose to study.
The idea of project-based learning is not a new one. Indeed, it is far older and more ubiquitous than the invention of subject matter based courses or classes. In the USA, we often have historically tried to balance a public education that makes for “well-rounded citizens” with an education that helps ready people to “earn a living.” More recently, we seem to be focused only on the latter goal. In addition, we now seem to believe it is okay for people to go into great debt in order to secure an education. Putting resources into educating the next generation however, is not something meant to benefit only that next generation, but all generations to come.
Rest assured, it is not only Ph.D.’s in statistics that have challenges addressing problems in multiple ways. As Norton Juster in The Phantom Tollbooth suggests, many of us are prone to “jumping to conclusions.”
Precisely because we humans have such an exciting and completely new set of opportunities, challenges and dangers facing us now, it is more vital than ever to be flexible in our approach to problems. Under pressure, people are prone to fixate on the first approach even more than they usually are. How can we possibly believe this is a good time to cut back on public education? We need a citizenry who are not only knowledgeable but versed in a variety of ways to problem solve. It certainly won’t be enough to know what answers others have given to problems in the past. Why? Because they will be facing literally unheard of conditions. We need to let them at least jump to a different set of conclusions than the previous generation. Hopefully, they’ll do even better than that and not jump to conclusions at all. Rather they will work in cooperative groups to solve complex novel problems using the skills and confidence that were built at Horizons U.