If you ask one scientist to guess how many jelly beans are in a jar, unless they have worked specifically on that problem before, their guess won't be very accurate. But if you ask 500 random people, the mean of their responses will be quite accurate.
If you ask experts to predict the future of science and technology, will they be more accurate? SciCast
, the government crowd-sourcing project, hopes so. They are asking for participants to make their predictions.