> The whole investigation took me 5 minutes. A Web search, a few clicks, and some ctrl+f‘ing I think this is the core issue with this post. There is an inters

> The whole investigation took me 5 minutes. A Web search, a few clicks, and some ctrl+f‘ing I think this is the core issue with this post. There is an intersection between how much this new claim would update my mental model and how much time it will take to prove. The idea of "how much time it will take to prove" is important, because we all have a limited amount of intellectual bandwidth that we use to receive and integrate new information. Intellectual bandwidth means we need prioritization. Without prioritization we have two suboptimal states that are available. The first is where we are being DDOS'd in bad-faith by low-value claims, that have some amount of work to prove. The second is that whichever claims we do resolve are likely not the set of claims which result in optimal utility; updating our mental model results in utility, assuming that our update function only accepts "good" updates. One strategy, that I think I use, is to break things into four categories: (1) A claim close to my current mental model (2) A claim not far but not close to my mental model (3) A claim far from my mental model (4) A claim unrelated to my mental model. If a claim is close to my mental model (1), I'd probably just accept it out of hand. This newspaper clipping is one example, because proving that the clipping is real does not update my mental model (Nor does disproving it, honestly), so it is low value. This is the cheap filter that reduces redundancy. If a claim is far from my mental model (3), I'll probably just dismiss it out of hand. I'm not going to go to the work to prove or disprove something, insofar. There are exceptions here, wherein the statement is so orthogonal to the mental model that it might have high value and is therefore still with looking into. In my experience, claims of high value that are completely orthogonal from my mental model tend to be rarer than claims of low value. One case I can think of is an article on oil prices I read, where there was a claim that the breakeven oil price was ~$200 a barrel as support for Iran having a broken economy. This didn't make sense, as it would mean Iran is not a viable economy even with oil, but then why are western nations wasting political capital embargoing an unviable product? So I spent time investigating it (Investing intellectual bandwidth). In the end, it seemed like a bad-faith argument as it assumed that the oil price would be the only thing to balance the Iranian national budget and didn't speak to the fact that Iran is subject to artificial export constraints via embargoes. Engaging with claims in this category seems to result in finding a lot of gravel and maybe a piece of gold from time to time. The medium category (2) is where I think most of us should be spending our time, the frontiers of our mental model where things are both foreign as well as familiar. When you can take different parts of evidence and see that some of it makes sense and some of it doesn't. I think this area is ripe for bad-faith disruption, as I'm much more willing to engage with this type of premise. I remember reading and putting down a book lately because it proposed an interesting question, around what has changed in the America of my parent's and of today. It was asserting, without even worrying about supporting evidence, that traditional family values were the cause. The last category (4) is not an unimportant one, which is just the area where you might say "I don't know" or "Why does that matter". If someone claims Charles II was born in 1629, I also wouldn't derive much value by looking it up and saying "No, he was actually born in 1630". In that sense, it is low-value information because it doesn't intersect with meaning for me. Overall, I think the problem with the line of thinking I quoted is that it puts a lot of burden on the individual for what usually has zero or negative value to that individual. That work exists whether or not a URI is included, and is possible exacerbated by there now being "evidence" provided even insofar as the linked context provides no value. Aside: The phrase "ctrl+f'ing" written out is a bit humorous.