Bots and Simulations > Simulation Emporium

The case of blind Mr Magoo

<< < (2/3) > >>

Endy:
As long as it's not something like the One or one of your complex vision schemes, I do okay.

This was particullary hard since I couldn't use any of the standard stuff. With no sight it was about impossible to tell what was being hit or shot at.

Sometime I'd like to try and make a truly combined senses bot, I think the more stuff we give them at the start the better their overall evolution goes. Just need to balance the complexity with safety and simplicity. The One does remarkable well in this regard giving rise to fairly complex mutants.

I do like the fact that we can actually mix symbols and numbers like this. I don't think I've heard of any high level programming language where you can switch from symbol to number and back again, so easily. I wish the others did have the capability, we could combine conditions and operations in one line. :)

Testlund:

--- Quote ---Mister Magoo is a curious bot - it does not use eyes at all.
--- End quote ---
That's how a bot should be! It's not realistic with single cellular organisms with eyes. I just wonder if I can try this one in 2.4.A. It's not possible to run a successful simulation in 2.4.A with those settings you mensioned though.

Numsgil:
Hmm, this seems familiar...

Again, Darwinbots are not necessarily single celled analogues of real life orgtanisms.  They are what they are, and as all adaptive complex systems, are subject to mutation and natural selection, which is what we're really interested in.

Single cells are just the closest natural analogues in real life, and so are what we tend to pull inspiration from.

Second, and perhaps more controversially, there are cells with eyes of some form or another.  I can't get alot of information on them online, but I do find numerous references to such organelles.

Google "eye organelles" and see what I mean.  As to the capabilities of these eyes, I have no idea.  I assume at the very least they can detect differences in light.  Probably not as complex as eyes in Darwinbots, but still.  So nyah :P

Endy:
Umm...actually that was one of my reasons for constructing this bot. I hoped that the sheer lack of eyes would spur it to develop them on it's own.

I personally just see the eyes as a sort of senseing organ(s). Unless some major work is done into computationally inexpensive field propagation(probably hardest thing to simulate), they're are the best we'll have for a long while.

Numsgil:
Basically, if you can sugest something radically better than eyes, I'd probably go with it.  Eyes are one of the main bottlenecks for the program.

But I just don't see what that would be.  Eyes seem to work well enough at the moment.

Navigation

[0] Message Index

[#] Next page

[*] Previous page