037 SMITH JOURNALOf course, there is one creature in nature that
already thinks this way. And it makes honey.
“A lot of people think the queen bee is like a
monarch, giving orders and telling the other
bees what to do,” says Professor Jean-Pierre
Scheerlinck, Director of the Centre for Animal
Biotechnology at the University of Melbourne.
“But it’s actually the reverse. Bees make
decisions in a much more democratic way.”
Scheerlinck has studied bee behaviour closely,
and he says there’s a lot we can learn from the
way a hive – made up of tens of thousands
of individual bees – ‘thinks’. Just look at
‘swarming’, the process whereby a hive splits
into two or more separate hives. Swarming
usually occurs when a hive gets too crowded,
so half the bees buzz o with a queen and
form a new hive. These swarms are often
faced with a sticky problem: where to go.
How do you fi nd the optimal location for
a new hive, one that won’t get raided by
predators of very little brain?
“The swarm sends out scouts,” Scheerlinck
answers. “And the scouts go and check out
the forest. When they return to the colony
they do what’s known as a ‘waggle dance’,
dancing around in a fi gure of eight. How
fast and how long they waggle indicates
the strength of that potential hive site, and
the angle of their bodies is the angle between
the new hive and the sun. The guy who
fi gured out that one got a Nobel Prize.”
Di erent bee scouts dance for their chosen
location, which causes more bees to go and
check out potential sites. These bees then
join the party, dancing for whichever hive
they think is superior, casting their vote in
very small body language. Soon the surface
secret power of groups isn’t agreement;
it’s argument. “We’re not trying to build
a system that will get everyone to agree
together. Our system depends upon lots
of di erent people passionately holding
di erent perspectives.”What’s so great about consensus anyway?
Lynch mobs have consensus. Divergent
thinking is far more interesting. Remember
Galton’s ox: some of the crowd guessed high,
others guessed low. The correct answer was
the average of both. (Interestingly, research
into information cascades suggests the
crowd probably would’ve been less accurate
if everyone had known each other’s guesses.
That’s crowds for you.)While SWARM has been funded to build
better security intel, van Gelder says a hive-
style ‘arguwiki’ could be the key to solving
all sorts of real world problems. Imagine a
real-time forum where average people could
crowdsource the solutions to climate change,
or come up with more ei cient and equitable
healthcare. Perhaps make better honey. You
could remove the frailties and fi ckle biases of
individual reasoning and tap into some sort
of objective, unifi ed intelligence. A whole planet
of digital bees, waggle dancing. It’s pretty much
the Earth envisioned in Hitchhiker’s Guide To
The Galaxy: a computer so vast that life itself
forms part of the hardware. A machine that
doesn’t just calculate answers: it can learn
to pose better questions.“We’re drawing together what philosophers
and mathematicians have known for some
time,” van Gelder says. “Good reasoning
is our best chance at reaching the truth
of a matter. And ultimately, this is all
about fi nding the truth.” •of the colony is waggling with competing
arguments. “After a while, consensus is
built,” Scheerlinck says. “The bees decide,
‘That’s enough data.’ And they make a
decision as a whole and go colonise this new
spot.” Biologists have tested this process, and
it seems to work. Researchers from Cornell
University presented swarming colonies with
fi ve identical hives. Four of the hives were
slightly smaller. Once the scouts found the
larger, fi fth hive (let’s call it the Penthouse
Hive), more and more bees began checking
it out. Eventually, the colonies reached a
consensus. They all chose the Penthouse.You may have gathered from the name, but
this ‘hive mind intelligence’ was part of the
original inspiration behind the SWARM
Project. “Swarming is a pretty good analogy,”
van Gelder explains. “It’s essentially what
happens on our platform. People put up
di erent ideas, they put something forward
into the common space, and they say to
everyone else, ‘What do you think of this?’
There’s comment and feedback, and then
the group collectively makes up its mind.
They vote on the strongest reasoning.”
He says it’s the di erence of opinions that’s
crucial here. Just like a swarming hive, the