I just completed, with 100% success, the certification for the California Civil Rights Department. This is the third such sexual harassment in the workplace test I have taken in the past three years. It is a function of having changed employers three times. In some ways, I feel that taking tests and certifications are insults to my intelligence. I don’t think I’ll ever be free of that feeling.
I cannot remember who coined the term ‘prediction is compression’ but it makes very good sense in this context. Watching the instructional videos which played out various scenarios of inappropriate behavior was in fact very cringe inducing. The course had a trigger warning, which I ignored, but that didn’t change the fact that I wanted the victims of these untoward aggressions to throw a punch in medias res. But it was not to be, and I had to suffer through it all. At the end of each video were simple statements for which I had to answer true or false. This was an excellent course in that it is almost inconceivable to fail it, certainly by design. It digested the complexities of the law down to about 20 true or false questions. I am now certifiably aware. Yet there is no way I could explain the law, or the series of emotions of bearing through the course. In the end, I have a certificate which is a compressed object testifying to a large number of facts.
The devil is in the details.
And yet predictions can be made. This week saw the death of Jim Simons, a man who has been in the periphery of my mind since I read a biography of him over a decade ago. Confidentially, I always desired to be a Wall Street ‘rocket scientist’, but never quite managed it. I came close to it a couple times but never successfully made the jumps or the connections. I decided thus to play in the second string world of corporate business intelligence and enterprise software, but have always admired the likes of Emanuel Derman, Walter Wriston, Ray Dalio and Nassim Taleb. Margin Call is one of my favorite movies, then again so is The Accountant. The point is that some people can personify the accurate prediction of complex, dynamic adaptive systems, and that is the godlike skill that we can sometimes witness in genius. The wisdom of these people is compressed. They make connections in multidimensional realms and compress them down to heuristics and disciplines made manifest in the actions of their companies, processes and algorithms.
We can expect compression in many fields. We can expect, if this revolution of artificial intelligence proves fruitful in multiple domains, compression of all sorts of details. It also means, with any luck, that the details can be decompressed. That the implications of the true/false questions can lead us back the the principles of the laws. And if and when we leave the geniuses in the loop without automating everything, then we can build protocols which will be efficient and complete. We may test the very limits of Gödel’s Incompleteness. Why? Because we will incorporate two kinds of intelligence. We will have both lossy and lossless compression. We can optimize the combination of the two - sometimes we can use an aphorism, sometimes we can use the capture of a full-cringe experience on video, even in virtual reality and impress things on our minds and psyches in the way motion pictures were originally conceived as an educational tool.
Artificial intelligence systems must therefore augment without automation. You therefore don’t automate people out of the process, you enable them to learn and think during the process as part of the process’ own security. You need both types of intelligence.
Who can predict? All of us. But as Neils Bohr said:
Prediction can be very difficult, especially about the future.