If you look at social
species, they always tend to work better and more efficiently whilst in large
numbers. This is the reason why certain creatures, such as bees, flock together
- it allows them to react in favourable ways by combining all the information
the group has as a whole, rather than each member's individual knowledge.
Researchers are now applying this approach to AI. However, can humans better
use collective intelligence in a similar manner? Louis Rosenberg, who manages a
Silicon Valley startup called Unanimous AI, believes they can.
The company has built a ‘hiveminds' platform that supports human
decision-making by crowd-sourcing opinions online. It is described as a human
swarming, a real-time method for parallel-distributed intelligence. Breaking that
down, it means that inputs come from real humans, where the computation is done
as it happens with all inputs coming in simultaneously. Rosenberg's platform
has, in a series of recent tests, made numerous predictions that are
surprisingly accurate, including the winners of the 2015 Oscars and the winners
of the 2016 National Hockey League's Stanley Cup for example.
These results aren't based on a single person's opinion but
instead derives an average that is gleaned from the wisdom of the crowd. Also, it
was observed that small swarms consistently outperform much larger crowds.
Also, the system indicated better conclusions when people picked an answer all
at once, because we tend to make better decisions as a group than as
individuals. Rosenberg believes that this hybrid human-computer approach, which
until now was either human based or all-out artificial intelligence, could
someday help us attain solutions to some of the world's most pressing
challenges. Rosenberg says, “Swarms will outperform votes and polls and surveys
because it's allowing the group to converge on the best answer, rather than
simply finding the average sentiment.” His team successes have led to them
receiving interest from a wide range of groups.
Source: DNA-17th-December-2016
