Artificial Intelligence is and continues to change the way we see things and the way we do things. For much of our lives there has been conventional wisdom, there are strong elements why conventional wisdom has lasted, but is it the best approach? In a recent book written by Nick Polson and James Scott titled AIQ published by St. Martin’s Press, NY, 2018, the authors examine artificial intelligence – what it is, how it works, where it comes from and how to harness its power for a better world. Note the optimism in the authors expectation, AI can help make the world a better place.

Within the book are some great and compelling stories which lead to the conviction, AI although there should be and will be concerns, on balance it is good for society. The issue remains how to interpret data for solutions. For generations and generations or through out history, governments and corporations have collected information. What they do with most of the information has been limited, now it is possible and probable there will be interconnectivity of the information. This can be very good for society.

To begin with you need to learn and understand a key concept called personalization or conditional probability. In math, conditional probability is the chance that one thing happens, given that some other event has already happened. The classic example is the weather forecast, if the forecast on your phone says rain 60%. Data scientists write this as P(rain this afternoon | cloud this morning) = 60%. P is probability;  the bar | in the brackets means given or conditional upon. To the left of the | is the event we are interest in. To the right of the | is our knowledge, also called conditioning event or what we believe or assume to be true.

Personalization run on conditional probabilities, all of which must be estimated from massive data sets in which you are the conditioning event.

The core idea behind personalization came from Abraham Wald who worked in Columbia’s Statistical Research Group and one of the clients was the military’s Office of Scientific Research and Development. Wald came up with a solution known as sequential sampling which show how factories could produce fewer defective tanks and planes just by implementing smarter inspection protocols.

Another project Wald worked on was devising personalized survivability recommendations for aircraft. In WW II, the military sent planes to drop bombs on the other side, many of them did not return. How do you improve the chances more would come back? Wald came up with an algorithm to improve the survivability of any model plane using data on combat damage. His task knowing the conditional probability that a plane has taken damage on the fuselage, given that it returns safely. The question to be answered was what is the conditional probability that a plane returns safely given that is has taken damage on the fuselage?

It is important to note conditional probabilities are not symmetric. It is very important to be clear about which event is on the left side of the bar, and which side is on the right side of the bar.

Wald needed to estimate how many planes had damage to the fuselage and never made it home? Wald’s model after examining most of the possibilities of how a plane could be shot down to have reasonably accurate assumptions solved the problem from any of the navy’s planes. Conditional probability plus careful modeling of missing data proved to be a lifesaving combination.

Linking to dividend producing stocks, there will always be information you do not know, it is possible to do analysis to estimate, but there are numbers of the company you do not know. Fortunately there are things we do know, in general profitable companies will trade at higher multiples than non profitable companies; if there is a downturn in the economy dividend producing companies lose less than the general market and bounce back sooner. If you pick stocks from dividend companies your total return will often beat the market index.

There are more questions than answers, till the next time – to raising questions.