The Timing of Expert Predictions
Listening to experts is tricky business.
Hypothesis: When it comes to new technologies and other important changes, once the man on the street has a working understanding of it (not just an awareness of its existence), it is very unlikely that the domain expert telling us that it’ll never work will be correct.
Here are two interestingly related examples:
Economist and future Nobel Prize winner (2008) Paul Krugman wrote in 1998, “The growth of the Internet will slow drastically, as the flaw in ‘Metcalfe’s law' becomes apparent: most people have nothing to say to each other! By 2005, it will become clear that the Internet’s impact on the economy has been no greater than the fax machine’s”.
Economist and future Nobel Prize winner (2024) Daron Acemoglu wrote in early 2024, “Using existing estimates on exposure to AI and productivity improvements at the task level, these macroeconomic effects appear nontrivial but modest—no more than a 0.66% increase in total factor productivity (TFP) over 10 years. The paper then argues that even these estimates could be exaggerated, because early evidence is from easy-to-learn tasks, whereas some of the future effects will come from hard-to-learn tasks, where there are many context-dependent factors affecting decision-making and no objective outcome measures from which to learn successful performance. Consequently, predicted TFP gains over the next 10 years are even more modest and are predicted to be less than 0.53%.”1
Krugman was wrong. I would argue that so too will Acemoglu be proved incorrect. Some of this might already be obvious, as Maxwell Tabarrok points out. Regardless, as I’ll argue below, I’m not sure we should be very enamored with claims like Acemoglu is making.
Corollary: Once the domain expert recognizes that something is not long for this world (an institution, practice, industry, technology, etc.), it will last much longer than he thinks it can but most likely he will be correct that it is coming to an end.
Yet there is a limit on this as the accuracy of the prediction is inversely proportional to the scope of the prediction. Predicting a company will fail is much more likely to be correct (and will come sooner) than the same prediction for an industry. Predicting a society will fall is orders of magnitude less likely and, even if true, further off—generally to the point of being very unlikely to be connected to the basis for the original prediction even when it does come true. For example, anyone around the time of Christ predicting the fall of the Roman Empire was almost certainly wrong in the cause, timing, and chain of events even though they were superficially right in the end—Rome did fall.
Related to this is how pessimism (cynicism) sounds wise and optimism sounds naive if not foolish. Yet while pessimists may keep us safe, optimists build the world. Along these lines are words of wisdom attributed to George Bernard Shaw, “Both optimists and pessimists contribute to society. The optimist invents the aeroplane, the pessimist the parachute.”
Consider this 2x2 framework:
Notice a subtle distinction between the upper-left quadrant and the lower-right. Most new ideas fail including ones that will eventually succeed. Predicting this with a seemingly accurate track record is not difficult. Still, the pessimists can pound their chests. Once we are in the long run, we are looking backwards almost by definition. Therefore, the optimists' predictions now have the rare hinderance of hindsight bias. Everybody falsely remembers a narrative that leads to the reality we actually live in.
Similarly there is a difference between the lower-left and upper-right. I believe pessimists get charitable credit to be point of undeserved excuse for when, as is often the case, their predictions fail in the long run. Optimists on the other hand receive less than their full due as their correct predictions are too surprising for them to have completely foreseen.
Pessimism as a disposition suffers from the dual aliment of being right in unimpressive ways and wrong in hidden but costly senses. Optimism suffers from an inability to be correct when people are looking and a lack of respect for being mostly correct in the important aspects after most have ceased to care.2
Examples for the original hypothesis and its corollary:
New: telephones, fax machines, the Internet and all of its subdomains, Bitcoin, et al., AI, ...
Dying: railroads, ICE vehicles, newspapers, radio and television as we’ve known them, each iteration of music hardware media3, ...
Two wonderful and related sources on this are The Pessimists Archive Newsletter and Human Progress.
I adore the figures extended out to the second decimal. I understand why the figures would extend out to this level of precision in a scientific paper of this nature with magnitudes that imply really big impacts of small changes. It is likely inescapable, but it is also a point of folly.
Pessimism is a broken clock. Optimism is Rodney Dangerfield.
Vinyl album, 8-track, cassette, compact disc, mp3 (purchased individually), etc.