These comments are oddly common, but I've never understood the value of simply repeating the definition of uncertainty words like "may." Who is being informed here?
Not every correlation deserves to be published in the press. If you think about it, it is weird that they even make articles out of statements that only "may" be true.
It is a matter of thresholds, of consequential bets over provisional non-trivial assessments, and of journalistic duty (taking note - the idea is not from the analyst but from sources; journalists take note that sources made statements).
Analysts work speculatively. And the results of the speculation are not trivial when supported by strong elements.
Errors are fine when they are due to a random process. When the press pick up claims because they are likely to attract readers it's a different story. Errors are not random, they are the result of an optimization process and are very unlikely to happen on the other side of the claim. (For example consider the overwhelming number of articles that have been published with the title "Scientist may have found life on mars" over the past 30 years)
For this reason, "may" in a press article, doesn't carry the same meaning regarding uncertainty than "may" in common language. Which is why I disagree with your comment:
> I've never understood the value of simply repeating the definition of uncertainty words like "may."
In the contrary, I think it is worth reminding people what may actually means here.
Now, I agree with what you and others have said about the usefulness of discussing uncertain claims, and I don't think there is a real disagreement here.
I am being informed about the public opinion. Which I value more than articles. (Don't come me at with AI, I'm well aware that it makes all human communication bad and pointless in theory and more and more in practice, thx.)
LastTrain · · focus · HN ↗
[dead]
schiffern · · focus · HN ↗
These comments are oddly common, but I've never understood the value of simply repeating the definition of uncertainty words like "may." Who is being informed here?
bnegreve · · focus · HN ↗
mdp2021 · · focus · HN ↗
Analysts work speculatively. And the results of the speculation are not trivial when supported by strong elements.
gilrain · · focus · HN ↗
I thought about it. And it's very stupid to imagine uncertainty should be off limits. I hope that isn't what you meant.
bnegreve · · focus · HN ↗
schiffern · · focus · HN ↗
I prefer error bars be worn on the sleeve, so I find "may" on display to be refreshing, not dismaying. YMMV
What's your alternate explanation to explain the absence of fentanyl in Italy, and how can my country get so weirdly randomly lucky?
bnegreve · · focus · HN ↗
For this reason, "may" in a press article, doesn't carry the same meaning regarding uncertainty than "may" in common language. Which is why I disagree with your comment:
> I've never understood the value of simply repeating the definition of uncertainty words like "may."
In the contrary, I think it is worth reminding people what may actually means here.
Now, I agree with what you and others have said about the usefulness of discussing uncertain claims, and I don't think there is a real disagreement here.
bmacho · · focus · HN ↗
I am being informed about the public opinion. Which I value more than articles. (Don't come me at with AI, I'm well aware that it makes all human communication bad and pointless in theory and more and more in practice, thx.)