Vernor Vinge, the author/scientist who gave the Singularity its popular name, recently predicted that it will happen in the next couple decades: Quote:
In a new interview, Vernor Vinge predicts the Singularity within 21 years. 'I'd personally be surprised if it hadn't happened by 2030,' he announces, saying humankind may become 'the only animal that has figured out how to outsource its cognition' to superintelligent machines.
In the new film Transcendent Man, Ray Kurzweil predicts that it will occur by 2045. Who is right? Only time will tell, but maybe these predictions are not really inconsistent. What if Vinge is predicting when the Singularity will most likely occur while Kurzweil is predicting the expected (50% probability) time, i.e., the mode and median respectively? For a log-normal distribution these can be very different values depending on the variance.
Assuming the mode and median are 21 and 36 years respectively, I calculated an EV (mu) and standard deviation (sigma) of 3.58 and 0.73. The resulting distribution is graphed below with the red line plotting the PDF and the blue line plotting the cumulative probability. Notice the peak of the red log-normal distribution is at 2030 (Vinge's prediction) and the blue line crosses 50% at 2045 (Kurzweil's prediction). Also note that probability of the Singularity occurring before 2030 is only 23% while the probability of it happening before 2100 is 90%.
Shane Legg predicts human-level AI most likely by 2025 and with a median (expected) date of 2028. Using the same technique as I used above I calculate mu=2.89 and sigma=0.427 yielding this prediction curve: