Robots evolve altruism
news.sciencemag.org
news.sciencemag.org
I had trouble reading beyond that line, since it's simply false. A quick reading of "The Selfish Gene" by Dawkins would correct that impression.
Imagine if a sentence in a news story read, "For centuries, doctors have struggled to determine whether diseases are caused by an imbalance of the body's humours or by germs and genetics."
Would you defend that by saying, "It's not false at all. Germ theory didn't gain much currency until the mid-1800s, and we didn't understand genetics until a century later. So even if everyone immediately accepted germ theory upon reading it (which they didn't), that sentence is still an accurate history of doctors' opinions over the years"?
Reason #1: future favors, tit-for-tat types of responses. If I help a friend out in a time of need, some day when I am in need the same friend might be able to help me. Neither of us are related, but we've both increased our survivability. This effect becomes more prominent with communities, even communities of unrelated individuals: you all survive or die as a group.
Reason #2: a fluke of the system. Evolution doesn't provide perfect solutions, just "good enough" solutions (evolution is dumb). Thus a species that is very good at being altruistic for close relatives (a beneficial thing), but occasionally is altruistic for unrelated strangers (possibly not beneficial) might be, on the whole, better than one that is never altruistic.
This runs all the way from courageous knights errant to spotty teenagers feigning an interest in Habitat for Humanity.
So there actually are benefits to using robots instead of doing a simulation. Simulations are usually faster and cheaper to create but robots give a more realistic result.
The researchers then created virtual representations of
these robots on a computer so that they could observe the
robots' evolution over time.Dreaming a sympathetic wish,
As the lights blink faster and brighter...