Invalid.Inference

What is it? Why is it important?

     When I was in college I found myself in classrooms taking a LOT of classes, that never did seem to align itself with whatever “major” I had aligned myself with. People talk about things being “fluid” now a days, but it is indeed the best way to describe much of my college career. I honestly cant figure out exactly how students end up paying for college now a days, at least not without having a 40 hour a week job, that lets you travel to Europe for no particular reason and do it with luggage that cost more than my first car. My son, who is a Jr. now, suggested that that's what all the loans their complaining about now. Now I can see how it takes them so long to event try and pay back.

     This was all the exact opposite of how I did it. Pay for myself, work always and live extremely cheaply. Oh and take advantage of the idea that ran colleges and universities at that time, as long as your paying, you can take whatever classes you wanted to learn about, given 2 ideas: 1. You had to have all the preceding courses required 2. You had to score well enough in the class to encourage the professor to want you in the class. So I had, by the time I graduated, 6000 level math theory courses, listed as elective. But I enjoyed it and I learned a lot.

     One of the biggest things I learned was regardless how much I wanted it to work, certain applications of 'logic' just didn't work, we called it “Linear Logic”. If you took a course called “Logic” in college you know that it mostly consists of Venn Diagrams, and how the overlapping percentages add up to anything, or in some cases, everything. If you took Algebra, and understood it, in college you remember the single direct order that everyone did their best to adhere to, but one or two of your fellow students couldn't ever follow: “Show your WORK” work the problem backwards

     The majority of people on the planet start on step 1 and then work their way step by step, all the way through, to the last step, thus coming up with the right answer. Then there are folks like Charlie and I, who look at a problem, and either get it, or we don't, and IF we get it, we cant “show our work” because there was no step by step work laid out. Sometimes when I knew the professor would grade the work and give you “partial credit expansion” to help any points you lost for not getting the right answer. I would memorize the formulas, and then work the problems backwards. This “showing my work.”

     If you took any “extension math classes” Theory of Algebra, for example, this came in very handy. Your job was to take someone else s work and “work it backwards: to prove their theorem. Its one the great ways math moves ahead during certain times of scientific advancement. For example, when the Manhattan Project was in fill swing, two of their top scientists, Einstein and Oppenheimer created a safe process that would allow nuclear “energy to be released all at once, and helping to end WWII. Einstein specifically never wanted to create a bomb , he was working on the project, in order to harness that much energy as “power” that could be used to light up everyone's living room!

     Unfortunately, while Einstein and Oppenheimer, and another scientist you may have never heard of, Richard Feynman, were literally the only 3 scientists that could look at the issue and understand the problem. All 3 of them could describe it, talk about it, understand each other, but the one thing they couldn't do, was explain it well enough for anyone else to understand. They needed a way to explain this “problem” so they created a term called “Linear Logic”, also known as “straight line thinking.

     As the years passed this term got used more and more on “college campuses” by professors the trying to teach “new” subjects to the average High School graduate. When education and grades crossed paths they would explain that although “linear logic” is a valid way of thinking, it was not always 100% reliable. There would be an issue where a math problem is thrown up in front of the class that has a line that says “X+Y=2”, X=1 , and no other parameters or limits to the question..

     Looks like a simple question. Thinking in :linear Logic the answer would obviously be Y=1. But if you start throwing in some special circumstances,

 for example  Y=logarithmic progression of Z. The entire circumstances of the question changes. Meaning that in this case, NO 1+1 will NOT = 2.

     How does that play into the world of OI (Osteogenesis Imperfecta) research and the ideas about what ELSE it may encompass.

     For example, one thing that many OI patients, and Physicians are surprised by is that research , experiments, papers etc all point to a direct connection between cardiovascular issues and OI. As of this date however there is NO direct or biological connection between OI and many of the Cardiac issues that it seems to be involved in.

     High Blood pressure for example. In a national survey recently displayed the information that, of the respondents, 98% of the patients with OI, regardless of the type, had been diagnosed as having high blood pressure. There are some theories being bandied about among researchers . One suggests that the lack of quality collagen, the major ingredient that makes UP the pipes of the cardio vascular system, having high blood pressure is the LEAST of the issues that should be expected. But they still have not determined in a precise REPEATABLE way, exactly what the issue is that gets created and causes the BP to raise. Then the research group that's a tad younger than most, and has a system of learning more based upon the humans ability, or inability to process past physical accidents. Thus they are tell us that it is our capability or in capability to deal with physical trauma f living our life, is what causes the BP to spike high. NOT my favorite theory, by the way.

In the past we could have used linear logic to describe why this seems to be the path to the higher BP, but it doesn't quite fit. Taking into account all the changes made in languages in the last 10 years, what with the advent and use of AI, scientists needed a way to explain what they were working towards and what their “AI”s are expected to do. The term finally settled on was “inference”.

     Heading over to Webster.com we find that INFERENCE: “the act or process of reaching a conclusion about something from known facts or evidence “ So the computer scientists who are training these AI databases to reach conclusions in the most “human” way they can are really teaching them to to reach an INFERENCE. Therefore if using that definition you are reaching a conclusion using “KNOWN” facts and “evidence”, but its not actually correct. This then would be called an Invalid Inference. In some cases scientists have used this term to explain why or how their AI was having “Hallucinations”. Personally I've never liked applying that term to a machine program trying to solve a problem it was presented with, but in general usage it seems to fit rather well.

     Therefore an INVALID INFERENCE occurs when something looks like it should work, but doesn't. In this case its BIOLOGICAL processes that are happening naturally (OI) , but not logically. Therefore its an INVALID INFERENCE. I hope this clears things up, and I will do my best to stop using the term Straight Line Logic, or Linear Logic.?