why is Tetlock-style judgmental forecasting so popular within EA, but not that popular outside of it?
The replies so far seem to suggest that groups outside of EA (journalists, governments, etc) are doing a smaller quantity of forecasting (broadly defined) than EAs tend to.
This is likely correct but it is also the case that groups outside of EA (journalists, governments, etc) are doing different types of forecasting than EAs tend to. There is less “Tetlock-style judgmental” forecasting and more use of other tools such as horizon scanning, scenario planning, trend mapping, etc, etc.
(E.g. see the UK government Futures Toolkit, although note the UK government also has a more Tetlock-style Cosmic Bazaar)
So it also seems relevant to ask: why does EA focuses very heavily on “Tetlock-style judgmental forecasting”, rather than other forecasting techniques, relative to other groups?
I would be curious to hear people’s answers to this half of the question too. Will put my best guess below.
– –
My sense is that (relative to other futures tools) EA overrates “Tetlock-style judgmental” forecasting a lot and that the world underrates it a bit.
I think “Tetlock-style” forecasting is the most evidence based, easy to test and measure the value of, futures technique. This appeals to EAs who want everything to be measurable. Although it leads to it being somewhat undervalued by non-EAs who undervalue measurability.
I think the other techniques have been slowly developed over decades to be useful to decision makers. This appeals to decision makers who value being able to make good decisions and having useful tools. Although it leads to them being significantly undervalued by EA folk who tend to have less experience and a “reinvent the wheel” approach to good decision making to the extent that they often don’t even notice that other styles of forecasting and futures work even exist!
The replies so far seem to suggest that groups outside of EA (journalists, governments, etc) are doing a smaller quantity of forecasting (broadly defined) than EAs tend to.
This is likely correct but it is also the case that groups outside of EA (journalists, governments, etc) are doing different types of forecasting than EAs tend to. There is less “Tetlock-style judgmental” forecasting and more use of other tools such as horizon scanning, scenario planning, trend mapping, etc, etc.
(E.g. see the UK government Futures Toolkit, although note the UK government also has a more Tetlock-style Cosmic Bazaar)
So it also seems relevant to ask: why does EA focuses very heavily on “Tetlock-style judgmental forecasting”, rather than other forecasting techniques, relative to other groups?
I would be curious to hear people’s answers to this half of the question too. Will put my best guess below.
– –
My sense is that (relative to other futures tools) EA overrates “Tetlock-style judgmental” forecasting a lot and that the world underrates it a bit.
I think “Tetlock-style” forecasting is the most evidence based, easy to test and measure the value of, futures technique. This appeals to EAs who want everything to be measurable. Although it leads to it being somewhat undervalued by non-EAs who undervalue measurability.
I think the other techniques have been slowly developed over decades to be useful to decision makers. This appeals to decision makers who value being able to make good decisions and having useful tools. Although it leads to them being significantly undervalued by EA folk who tend to have less experience and a “reinvent the wheel” approach to good decision making to the extent that they often don’t even notice that other styles of forecasting and futures work even exist!