And then there's things that everyone knows and that are wrong. Like famous movie quotes. "I am your father, Luke". Known as the Mandela effect [1].
A fast-mode consensus has a high probability to create lasting wrong "truths", that will be difficult to dispell later.
The proper way would be to attach probabilities to information, but that would be too much for most readers, and impossible to do for most normal jounalists.
> The proper way would be to attach probabilities to information, but that would be too much for most readers, and impossible to do for most normal jounalists.
some scientific reporting takes the approach of defining certain words to represent confidence (probability) ranges: “we are weakly confident”… “moderately confident”… etc. they often explicitly define the range, too. that’s the case e.g. in IPCC reports, which is technical material that way more people i know than normal read.
i don’t think qualifiers like ”weakly”/“strongly” get in the way, but they do show just how uncertain most effects actually are. people don’t always share information out of altruism. frequently, information is shared in an attempt to persuade. and so there’s selective pressure for writing which makes a situation seem more black/white.
if you want quality communication at scale i’m not sure if your bigger priority would be introducing probabilities, or rather aligning everyone using the communication channel to value truth. Wikipedia does a far better job at presenting good information than most of the press, despite lacking probabilities.
> i don’t think qualifiers like ”weakly”/“strongly” get in the way, but they do show just how uncertain most effects actually are.
"Most" seems to caused by selection bias.
There is new areas where knowlegde has to be established first. Theories can be created, and experiments defined to confirm or reject them. But it takes time. That's the realm of fast tracking results the original article is about. The temporary results will be replaced later. Covid is such a case (even if ethics may get in the way of some experiments). "Most" of science is like this.
Then there's areas that cannot be verified by experiment. Probabilities are useful there, too, even more so, because they will persist long-term, even if there is consensus. Main examples are climate (no control group of earths to conduct experiments on) and cosmology (no control group of universes).
A fast-mode consensus has a high probability to create lasting wrong "truths", that will be difficult to dispell later.
The proper way would be to attach probabilities to information, but that would be too much for most readers, and impossible to do for most normal jounalists.
We're doomed in any case.
[1] https://en.m.wikipedia.org/wiki/False_memory#Mandela_Effect