June 2019 - Core Group Meetings

For the June 2 meeting at 7:00 pm EST.

A. Scott

1. Comments, observations, ideas, etc.​

2. Grants for Candle​

3. Candle description (started with Jack earlier this week)​

B. Jack

1. I'm now focusing on a single task: How to measure the impact of patterns of deception, or conclude we don't know how to do that and should fall back on Deception Count Density.​

This week I tried a new strategy: Creating glass box models of how specific patterns worked. This was very productive. We now have a simple model. It's immature and wrong, but it's a start.​

Late in the week I shifted to studying the literature for ideas. None found. But it has been useful in getting me into thinking about the problem in slightly new ways.​

Saturday morning I moved to the next phase, trying to create Deception Theory 1 (DT1) that has equations based on the ideas in the glass box models. This too is immature and wrong, but it's much more complete than the glass box models. Here are three interesting paragraphs from my working notes:​
The theory works on deception or truth because both involve belief, reasoning, inputs, etc. The theory is a highly simplified model of how the synaptic structure of the brain works to make decisions.
A good theory explains and predicts. The purpose of DT1 is to explain how patterns of deception work, explain how the average impact of individual patterns can be measured, and predict how truth literacy can be raised to reduce susceptibility to deception. The second explanation allows measuring the credibility rating of sources.
The core concept of DT1 is that a person’s decisions arise from the current state of their web of beliefs. Deceptions (false beliefs) and truths (true beliefs) form a web of many beliefs, connected by one belief becomes a factor in another belief. Planting a false belief in the web has the ripple effect of causing other beliefs to change in a false direction. The same holds for planting a true belief. While we talk of false and true beliefs, each is a belief whose truth level varies from zero to 100%.
My next step, when fresh, is to plug in actual cases of patterns of deception into the theory equations. That will show where it needs revision or what its limitations are. I find this type of synthesis so difficult that I can only do it in about the first two hours of each day, when super fresh. The rest of the day I work on less taxing work.​

C. Montse

1. I studied together with Jack his new glass box model and took several notes based on that. I also tested the model myself using an example identical to one he also picked, which was a good exercise to compare our results, which indeed turned out to be different because we focused on different bits of information. My overall conclusion is that this seems to be a very powerful approach because we would fully understand what a pattern does, and therefore why it has the effect it has. This is what medicine does when they study a virus, and they explain what it does and why we see the observed effects. However, precisely because this approach is superior, it is probably going to be harder.​

It was very insightful to see how to create a glass box model you have to actively seek to create one, it doesn't happen by default. Ultimately we do want a glass box model, but the current one still has some important flaws. If we can make it work elegantly, I would definitely go for this one.​

2. Studying the new glass box model, I noticed that if we take this approach, we would have to re-outline the paper. I made a revision, and I noticed that the modifications would be quite significant. If we come up with a strong glass box model to explain patterns of deception, then the paper would mainly be about presenting that theory. What we can do based on that (Source CRs, Article CRs, etc.) would be either on a second plane on that same paper, or on an entirely new one. If the theory is really groundbreaking, that is enough for publishing.​

3. Talking about grants, I started working on what is called a "Research Statement", a document in which your current and future research are presented. In general I think I will be needing something of that sort often in the near future, but specifically I was doing that because I found this call for applications, that seems to be right around my interests alley. Apparently they seem to be looking for very experienced researchers, but I could argue that I will be at an institution with very experienced researchers on that topic, that could guide or even collaborate with me. I think my chances to be selected are probably very low, but on the other hand it seems so specific that I don't think there's going to be too much competition, and at the very least it can be a good experience to learn about the process of grant applications. The deadline is on Thursday already, so I'll be working on that these days.​