[ExI] Google’s Go Victory Is Just a Glimpse of How Powerful AI Will Be

[Robin D Hanson]

On Feb 2, 2016, at 12:03 PM, Brian Atkins <brian at posthuman.com> wrote:
> A blog I read called Political Calculations uses dividend futures data to try and predict the market's upcoming behavior. Based on his work, the overall mass of market participants seem to have a very near term view, for example in this recent post he is discussing how most investors seem focused on 2016 Q1 and Q2 dividend data:
> 
> http://politicalcalculations.blogspot.com/2016/02/exploiting-future-for-s-500-in-week-4.html
> 
> So is it realistic to expect the market as a whole to react to something that, at best, is multiple years away?

There is a HUGE literature that looks for biases in financial market prices. Such biases are often found, but they are typically small, and if not typically go away fast once publicized. The main known bias is excess long term volatility. But that wouldn’t explain any particular mispricing. If there was huge consistent bias to ignore the future we’d know about it, and we don’t. 

> Perhaps your upcoming book will start to get the word out more. In the meantime I agree with Alejandro that investors focused more on specific stocks like GOOG and NVDA may be a bit more willing to invest based on longer term horizons, leading to the significant outperformance in those companies since October.

“since October” is WAY to short a time to see the effect of investing based on longer term horizons.

Robin Hanson rhanson at gmu.edu 
Future of Humanity Inst., Oxford University
Assoc. Prof. Economics, George Mason University
See my new book: http://ageofem.com