I noticed that talk also. Peter Norvig discussed the singularity in economic terms as well. He states that GDP growth is pretty flat over the last 50 years or so. However GDP growth over the last 100-200 years is very explosive. The problem with Kurzweil’s concept of the singularity is that it is independent of economic and social developments. His main metric are a few important, but selective, measures. His response to Norvig is a good one: GDP doesn’t measure deflationary tendencies well (or rather not at all). This is most relevant concerning computational power. However, scientific knowledge is hardly ‘measurably non-linear’. If one considers the output of scientist per person, it could be argued that there might some degree of saturation, due to the work that is needed of keeping track of other relevant work. Endogenous economic growth theory is hardly of much use, because it rests mostly of linear trends. Also output is most of the time measured in ‘hard assets’, whereas valuations of ‘soft assets’ is extremely hard. Furthermore, due to a globalised market, the growth of companies becomes more rapid. The dot-com bubble can be seen as this outburst of inherent uncertainty. However, the rise of the internet seems to be much a isolated event and there is no way to predict the nature of other such events and(!) that they occur at all. I think Thiel is right in pointing out that the notion of the singularity (events are unpredictable, progress is non-linear), encompassses a committment to the possibility that nothing might happen at all and that progress can lead to either catastrophic or ‘utopian’ outcomes. Certainly good speculators, such as Soros and Thiel himself, have a lot to say about dealing with uncertainty of this kind. The subprime debacle showed that sometimes it is worse to have a bad risk-model than having no risk-model at all. According to Taleb the black swan (an unforeseen event) is unpredictable in a certain way. The black swan is not the event that some model accounts with some probability, but something which fails the assumptions.