I found this article really intriguing as someone new to this field.
If I'm understanding this right, it seems like most approaches up to this point are focused on evolving a single "unit" or brain / person.
You have the concept of "nature" that selects which units will advance to the next generation and passing on "DNA". First based on fitness and now the new approach is novelty.
This may seem a bit naive, but has anyone explored adding a social dimension?
For the robot walking example, you have nature choosing novelty and those that made progress.
A basic social dimension could have units observing other units and sharing information.
But then there's a variety of other dimensions - a unit blocking another unit from walking (cheating), a bigger unit destroying a smaller one, maybe success via misc factors like physical symmetry / popularity.
Idk. Just curious to learn more and where the research is at.
If I'm understanding this right, it seems like most approaches up to this point are focused on evolving a single "unit" or brain / person.
You have the concept of "nature" that selects which units will advance to the next generation and passing on "DNA". First based on fitness and now the new approach is novelty.
This may seem a bit naive, but has anyone explored adding a social dimension?
For the robot walking example, you have nature choosing novelty and those that made progress.
A basic social dimension could have units observing other units and sharing information.
But then there's a variety of other dimensions - a unit blocking another unit from walking (cheating), a bigger unit destroying a smaller one, maybe success via misc factors like physical symmetry / popularity.
Idk. Just curious to learn more and where the research is at.