4 ms·

This analysis misses that most of what people spread is bullshit: things they intuitively feel or want to be true, but which are difficult or impossible to prov

This analysis misses that most of what people spread is bullshit: things they intuitively feel or want to be true, but which are difficult or impossible to prove or disprove. When bullshit spreads that authorities don't want spread, they censor. Neither party to the dynamic knows or cares if the bullshit is actually true or if there are compelling arguments either way. That would be too much work. Some day FB is going to censor some urban legend about zombies because it motivated some kid to kill another kid. Will you believe in zombies then?

AnthonyMouse 5y ago
You're just saying the same thing. Suppose the incompetent authorities aren't worse than random. They're just censoring whatever ambiguous things they don't like independent of their veracity. Then the expected chance that it's true is 50%. Coin flip. You add to that any amount of things censored intentionally because they know for sure that they're true and you're above 50%.
- civilized 5y ago
  The base rate of truth for stuff people spread is much, much lower than 50%.
  - AnthonyMouse 5y ago
    Which we know because we had a public debate and gave each side a chance to make their strongest case. Wait.
    - civilized 5y ago
      What debate do you want exactly? Two scientists at podiums yelling at each other? Nobody's stopping anyone from making the lab leak case and several people have published their best cases for it.
      - AnthonyMouse 5y ago
        Is this something you're putting forth as not having a probability around 50%?
        civilized 5y ago
        It could be, but censorship doesn't increase its probability. All it shows is the authorities don't like it. Your argument that they'd use facts if it were false is not relevant here because nobody knows which way the facts point.
        AnthonyMouse 5y ago
        This is actually a pretty good example. Suppose nobody knows, so the probability is ~50%. Now suppose there is any non-zero percent chance that they did know and the reason it was censored is that they didn't want to inculpate themselves. Then knowing that it was censored puts the percentage above 50%, right? It's a sliver of evidence in favor of the possibility that it could have been censored to hide the truth, in which case there was a truth to hide.
        civilized 5y ago
        I don't see how guessing that something was hidden because it was true is evidence that it is true. Guesswork isn't evidence.
        AnthonyMouse 5y ago
        Say that before there is any censorship, we start with a probability of 50%. People disagree, we don't know, it's plausible but not proven, that's our best guess. Then censorship occurs. This is new information. Evidence. If the censorship occurred because they don't like the speculation but they don't know any better than you, this doesn't affect the probability, because it's not dependent on whether the idea is true. If the censorship occurred because it was true and they wanted to suppress it, it was true. If there is any non-zero probability that this is what happened, the censorship is some evidence that it did, so the existence of the censorship increases the probability that it was true from 50% to some higher amount. If censorship only happened for the second reason then the calculation would be easy. If censorship happened, the probability that it was true would be 100%. It can happen for both reasons, so depending on how likely each reason is, the probability given censorship increases proportional to how likely that was to be the cause of the censorship. But unless the probability of the second alternative is zero, it increases by some.
- notriddle 5y ago
  > Then the expected chance that it's true is 50%. No it is not. Randomly generated ideas are much less likely than 50% to be true. For example, I could make two claims: COVID was a leaked research project, and COVID is a leaked research bioweapon. These claims cannot both be 50% probable, because the second claim is more specific version of the first and therefore must be less likely.
  - AnthonyMouse 5y ago
    Randomly generated ideas are gibberish. They don't even parse as a coherent sentence. For an idea to spread, it has to at least sound plausibly true. For it to spread to a large number of people in a forum open to public debate, it has to survive all of their attempts to convincingly refute it. The ideas that remain in contention are the ones that can neither be proved to be highly likely nor highly unlikely. In other words, the ones whose average probability is a coin flip. > These claims cannot both be 50% probable, because the second claim is more specific version of the first and therefore must be less likely. We're talking about average probabilities, not individual probabilities. Suppose the probability based on available information is 75% for the first and 25% for the second. Then the average is still 50%. When you put them in a box with a thousand other ideas with probabilities centered around 50% and pull one out at random, you're still at 50%. The things that didn't have a probability around 50%, i.e. the things that can be proved highly likely or highly unlikely, weren't in contention.
    - shadowgovt 5y ago
      > For an idea to spread, it has to at least sound plausibly true Given that just this week, QAnon believers showed up in Dealy plaza apparently expecting John F. Kennedy Jr would appear and re-appoint Donald Trump president, I think that assertion might call for a big [citation needed] sign hung from it.
      - AnthonyMouse 5y ago
        This doesn't apply to people who are mentally ill and/or trolling. It also requires public debate. This is why segregating people with stupid ideas from people who could correct them is a bad idea.

AnthonyMouse 5y ago

You're just saying the same thing. Suppose the incompetent authorities aren't worse than random. They're just censoring whatever ambiguous things they don't like independent of their veracity. Then the expected chance that it's true is 50%. Coin flip. You add to that any amount of things censored intentionally because they know for sure that they're true and you're above 50%.

civilized 5y ago
The base rate of truth for stuff people spread is much, much lower than 50%.
- AnthonyMouse 5y ago
  Which we know because we had a public debate and gave each side a chance to make their strongest case. Wait.
  - civilized 5y ago
    What debate do you want exactly? Two scientists at podiums yelling at each other? Nobody's stopping anyone from making the lab leak case and several people have published their best cases for it.
    - AnthonyMouse 5y ago
      Is this something you're putting forth as not having a probability around 50%?
      - civilized 5y ago
        It could be, but censorship doesn't increase its probability. All it shows is the authorities don't like it. Your argument that they'd use facts if it were false is not relevant here because nobody knows which way the facts point.
        AnthonyMouse 5y ago
        This is actually a pretty good example. Suppose nobody knows, so the probability is ~50%. Now suppose there is any non-zero percent chance that they did know and the reason it was censored is that they didn't want to inculpate themselves. Then knowing that it was censored puts the percentage above 50%, right? It's a sliver of evidence in favor of the possibility that it could have been censored to hide the truth, in which case there was a truth to hide.
        civilized 5y ago
        I don't see how guessing that something was hidden because it was true is evidence that it is true. Guesswork isn't evidence.
        AnthonyMouse 5y ago
        Say that before there is any censorship, we start with a probability of 50%. People disagree, we don't know, it's plausible but not proven, that's our best guess. Then censorship occurs. This is new information. Evidence. If the censorship occurred because they don't like the speculation but they don't know any better than you, this doesn't affect the probability, because it's not dependent on whether the idea is true. If the censorship occurred because it was true and they wanted to suppress it, it was true. If there is any non-zero probability that this is what happened, the censorship is some evidence that it did, so the existence of the censorship increases the probability that it was true from 50% to some higher amount. If censorship only happened for the second reason then the calculation would be easy. If censorship happened, the probability that it was true would be 100%. It can happen for both reasons, so depending on how likely each reason is, the probability given censorship increases proportional to how likely that was to be the cause of the censorship. But unless the probability of the second alternative is zero, it increases by some.
notriddle 5y ago
> Then the expected chance that it's true is 50%. No it is not. Randomly generated ideas are much less likely than 50% to be true. For example, I could make two claims: COVID was a leaked research project, and COVID is a leaked research bioweapon. These claims cannot both be 50% probable, because the second claim is more specific version of the first and therefore must be less likely.
- AnthonyMouse 5y ago
  Randomly generated ideas are gibberish. They don't even parse as a coherent sentence. For an idea to spread, it has to at least sound plausibly true. For it to spread to a large number of people in a forum open to public debate, it has to survive all of their attempts to convincingly refute it. The ideas that remain in contention are the ones that can neither be proved to be highly likely nor highly unlikely. In other words, the ones whose average probability is a coin flip. > These claims cannot both be 50% probable, because the second claim is more specific version of the first and therefore must be less likely. We're talking about average probabilities, not individual probabilities. Suppose the probability based on available information is 75% for the first and 25% for the second. Then the average is still 50%. When you put them in a box with a thousand other ideas with probabilities centered around 50% and pull one out at random, you're still at 50%. The things that didn't have a probability around 50%, i.e. the things that can be proved highly likely or highly unlikely, weren't in contention.
  - shadowgovt 5y ago
    > For an idea to spread, it has to at least sound plausibly true Given that just this week, QAnon believers showed up in Dealy plaza apparently expecting John F. Kennedy Jr would appear and re-appoint Donald Trump president, I think that assertion might call for a big [citation needed] sign hung from it.
    - AnthonyMouse 5y ago
      This doesn't apply to people who are mentally ill and/or trolling. It also requires public debate. This is why segregating people with stupid ideas from people who could correct them is a bad idea.