Dosage Matters

Glenn Harland Reynolds, aka InstaPundit, has a separate posting for musings and so on that is subscription only and comes as an email so we have no link. In the last one he titled the musing: Who Can You Trust? The sub-headline was: Did Robert Heinlein have the answer to deepfake photos and video? It was odd that he asked who in musings that included pictures and videos and even odder for a libertarian to consider Robert’s sci-fi book solution. We have two comments: First, we shouldn’t look for the gold standard in evidence and second, the dosage required to implement Robert’s solution is far too high.

Human reporting of events is often flawed because of the nature of human information processing. Thus, Glenn tells us:

In recent years, a common Internet refrain is “pics (or video) or it didn’t happen.”  Those were seen as proof.  And generally they were.

We will just use pictures to encompass stills and video. Well, no, pictures are not proof. Pictures always have have angles that include certain elements and exclude others. Both the left and the right has used pictures to argue that this event or that event was more or less violent. Then one side or the other can label an event as a peaceful protest, riot, or insurrection.

Sidebar: Death in Paradise has lasted into its twelfth season with many (most?) of the crucial plot devices centering on fooling pictures or clocks. Yes it is fictional but it is fun without being great and gives an idea of the problems with pictures. End Sidebar.

Pictures are helpful evidence but decision makers need a variety of sources of evidence including pictures, data (e.g., injuries and deaths, dollars of damage, and numbers involved), and personal reports. Rarely should any individual source be compelling to a decision maker. Rarely, if ever, does all the evidence lead to certainty.

Of course, many folks today aren’t looking to be decision makers. They have made their decision and looking for some evidence, however scant, to support that decision. Isn’t jury (and judge) selection for the trial(s) of The Donald going to be fun!

Our second concern is that Glenn, at least at a musing level, seems to take Robert’s creation of creatures called Fair Witnesses as an interesting idea. These creatures are:

[A]n institution that plays a minor role in Robert Heinlein’s Stranger in a Strange Land, the “Fair Witness.”  The Heinlein Society’s concordance describes a Fair Witness as a “Person rigorously trained to observe, remember, and report without prejudice, distortion, lapses in memory, or personal involvement.” [Emphasis added]

We have chosen to call Fair Witnesses creatures rather than humans because it is not clear that they are still human after the degree of rigor necessary to create them. Here is where the title, Dosage Matters, comes in. We already have such people. We call them policemen and related titles. Policemen are trained but the degree of rigor is not sufficient to be near one-hundred percent effective at reporting without prejudice, distortion, lapses in memory, or personal involvement.

To get anywhere near the goal, the rigor of the training must take the human out of human information processing. We think the price of Robert’s creation is too high for the possible benefits. And the benefits are limited because of the problem of getting the Fair Witness in the right place at the right time. It would be too kind to call a Fair Witness a government owned slave.

Police and related professions need more training to be more like a Fair Witness. At some point, however, any marginal improvement can only be called cruelty. Glenn, of all people, should have spotted that.

Skiing And Human Information Processing

We’re back from five days of skiing. Vermont was great and Killington was at its best midweek in March. There were zero lines and no crowded intersections. We got two nights of snow and the weather was nearly perfect for skiing. Two interesting accounting and economics issues that we noticed are that Killington has gone cashless and you can pay extra (but not cash) to cut in line. A little on them before we get to human information processing.

You can’t spend actual currency anywhere on the mountain. One interesting solution they have is machines that will turn currency into prepaid cards that Killington does accept. When you can’t spend currency you don’t end up with low denominations and by the end of the week tipping was a problem.

You can also buy a special ticket, we think it is called express, that gets you to the front of the line. Midweek in March the express ticket would be totally wasted as there were no lines. When the unavoidable survey came our way we told them so.

A personal highpoint was getting to visit with three friends from high school. As we have often told readers, human memory is notoriously unreliable despite humans being insistent it is. Back before we could drive, we had to find folks to take us skiing. John’s parents often did. Four of the five of us were on a trip that involved a vehicle accident. We all agreed that we had been to Loon Mountain and were on the Interstate when we hit an ice patch and eventually the guard rail. We all agreed that the police car came in and did a 360. We all agreed that we moved to stand behind the guard rail for safety. Two of us were insistent that John’s mother was driving. One was insistent, “I was in the front seat with John’s dad,” that they were wrong. The one of us that had studied human information processing stayed quiet on the driver issue. Since John and his parent are dead we are unlikely to ever know who was driving.

The point is that human memory is unreliable. Old memories are even more unreliable. The vividness of the memory doesn’t tell us much. Therefore when you hear folks talking about this or that changed their lives, you need to ask for actual evidence if they are trying to influence you. Otherwise, it is just a nice story and not really relevant to decision making.