[ExI] Convergence

[Anders Sandberg]

On 22/10/2012 17:19, David Lubkin wrote:
>> (We suspect you've noticed that all those things that were predicted 
>> to go
>> away are all still alive and well. So much for that prediction.)

Beautiful. Simply beautiful.

The big problem with predicting the future is that even when you get the 
analysis right, it might still take far longer than you expect for it to 
come true. I am reading a book on technology forecasting by Ayer from 
1970 that among other things mentions the failure of "radar owens" 
(microwaves), suggesting they were 'a flash in the pan'. No, it just 
took 15 years longer (and the introduction of the rotary plate) for them 
to become ubiqitious.

(Ayer also skewers using performance curve envelopes in a way that 
sounds perfectly applicable to Kurzweil critics today. All in all, a 
pretty good book - reading old futurology books is always helpful, since 
you can see what methods actually did make the right predictions and why. )



-- 
Anders Sandberg,
Future of Humanity Institute
Philosophy Faculty of Oxford University