Today was the first time I’ve heard Peter Thiel speak in-person and he did not disappoint!
I spent Saturday and Sunday at the 2009 Singularity Summit, held for the first time in New York. Michael Vassar, Singularity Institute’s president, did an exceptional job of assembling an all-star cast of speakers.
Talks covered a variety of subjects ranging from quantum computing to radical life extension to venture capital. For those unfamiliar with singularity, a brief description is here.
Peter starts by taking an audience survey, asking which of seven doomsday scenarios people are most frightened by.
The audience sees two clear winners:
a) runaway biotech, e.g. mutated diseases wipe out civilization or
b) that the singularity takes too long to happen (title of the talk).
Peter suggests that the recent credit/financial crisis actually has to do with singularity. Our society is on shaky ground: millions of people have money invested in pension funds (and the like) which they expect to produce returns of ~8% per annum, even despite various period of volatility. What people forget is that the past returns we’ve come to expect have been generated during periods of compounding innovation; cutting-edge tech has bolstered all other lines of business that rely on technological gains to increase productivity, etc. Innovation is a necessary component of a healthy, upward market. Innovation has stalled.
Peter recently talked to an LP (an endowment) who was thinking of re-allocating its portfolio to hold less than 5% in venture capital as its recent returns had not matched expectations. Peter told the LP that they’re thinking is flawed: if their 5% allocated to vc eventually does not return a mega profit, the remaining 95% allocation is going to be worthless anyway.
“Investing in technology as a VC is like Pascal’s bet: if that isn’t going to work, nothing will.”
Peter then asks: has recent technological progress really been slower than it should be? He cites as support the fact that a) wages have not increased in terms of real dollars since 1973 b) that California (the most innovate place in the world) can barely pay its bills and c) that VC’s haven’t seen good returns in 11 years.
The credit crisis was really a technology crisis. Credit really is a ‘claim on the future’ and credit only works with a background of growth. Peter then suggests that the credit crisis was really an offshoot of the dotcom crash. The dotcom crash in his opinion, occurred because the ˜innovation story’ was not happening as quickly as people expected. People panicked and bailed. When stocks (i.e. those necessary 8% returns) tanked, housing and leverage swooped in to make-up the difference.
Peter states clearly: ‘the boom-bust cycle is over’. There will now either be a long period of malaise, or we’ll see a great boom (he’s obviously betting on the former).
Peter slams the NASDAQ explaining that most listed companies are actually not technology companies, rather, they are zero-sum bets against technology (ex: investing in Google is feeding a monopoly on search, Microsoft doing well means linux suffers, etc). NASDAQ companies are more like banks than they are true innovators (Sun may be an exception).
The Madoff scandal occurred because people assumed that even despite a drought of innovation, an 8% annual return with no risk was still safe. This is crazy. People need to dramatically adjust their return expectations; only great risk will produce great returns.
Peter talks a bit about his fund and where he sees opportunities. Suggests that an under-invested area is optimizing the division of labour between human and computers: it doesn’t need to be all or nothing. He provides an example of how PayPal only managed to save the company from fraud by dividing up the fraud protection process using a combination of humans and computers.
Society currently suffers from a free-rider problem relating to science and technology. Everyone expects great innovation to manifest, but someone else will do it. Large companies have disincentive to think with a long time horizon; they need to make quarterly numbers. If China is to surpass the US (Thiel seems to think this may eventually) it will be due to this time dimension.
— WebMetricsGuru also has great coverage of the Summit here.