Donnadogsoth Posted October 1, 2015 Posted October 1, 2015 A friend years ago suggested bugs are "pure physics" without any consciousness at all. It occurred to me that the Turing test is aiming too high. It would be simpler to try to establish whether there is any meaningful difference between the mental processes of a biological ant, and of a robotic ant. How close are we to replicating ant consciousness as measured by complexity and precision of behaviour?
rosencrantz Posted October 1, 2015 Posted October 1, 2015 Russian scientists have created a device that enables them to control cockroaches
A4E Posted October 1, 2015 Posted October 1, 2015 I watched something long ago about a person studying ant behavior in extreme detail down to each ant by marking them, and the conclusion was that every single ant behaves differently. This was not a surprise to me, as I have always suspected that there is a lot more brain processes going on in insects than we think. How close are we to replicating ant consciousness as measured by complexity and precision of behavour? 100s of millions of years with survival of what works VS decades of programming I'd say not even 1% close.
Donnadogsoth Posted October 1, 2015 Author Posted October 1, 2015 I watched something long ago about a person studying ant behavior in extreme detail down to each ant by marking them, and the conclusion was that every single ant behaves differently. This was not a surprise to me, as I have always suspected that there is a lot more brain processes going on in insects than we think. 100s of millions of years with survival of what works VS decades of programming I'd say not even 1% close. Okay, then a followup question: How close are we to producing robot bugs that we can't tell apart from natural bugs? Presume naked eyes only and not allowing a squish test.
A4E Posted October 2, 2015 Posted October 2, 2015 Okay, then a followup question: How close are we to producing robot bugs that we can't tell apart from natural bugs? Presume naked eyes only and not allowing a squish test. The problem with trying to guess this is that there is probably a lot of R&D going on (and been going on) in this field outside of the public eye because insect lookalikes are very valuable for espionage, and a wide range of other uses.
ribuck Posted October 2, 2015 Posted October 2, 2015 How close are we to producing robot bugs that we can't tell apart from natural bugs? Bugs are relatively complex creatures. They need to be able to fly, evade predators, etc. Worms are much simpler. There's a type of roundworm with fewer than 1000 cells in its body (including just 95 muscle cells), and only 302 neurons in its brain. The OpenWorm project is a worldwide collaborative project to try to simulate this worm in software. You can read about it here: https://en.wikipedia.org/wiki/OpenWorm and you can join in if you want to be a part of some exciting emerging science.
Koroviev Posted October 2, 2015 Posted October 2, 2015 I think the other question is do we consider something that looks and acts like a bug to have "bug consciousness?" You can create a program that acts and reacts in certain ways based on certain circumstances, but at the end of the day it is still following a script that a person wrote. Along the same lines there is the rock, paper, scissors robot made in Tokyo that has a 100% win rate. It works by using a high speed camera to see what shape your hand will take then chooses the winning shape seemingly instantaneously. Does that make the robot really good at rock, paper, scissors or is it simply cheating? If I was allowed to choose based on my opponent's choice I'd be really good at rock, paper, scissors too http://www.k2.t.u-tokyo.ac.jp/fusion/Janken/index-e.html
Carl Green Posted October 2, 2015 Posted October 2, 2015 How could we even get to a point of saying that we know what ant consciousness is, in order to confirm if we've replicated it or not?
shirgall Posted October 2, 2015 Posted October 2, 2015 How could we even get to a point of saying that we know what ant consciousness is, in order to confirm if we've replicated it or not? Just thinking about it gives me hives.
Donnadogsoth Posted October 2, 2015 Author Posted October 2, 2015 How could we even get to a point of saying that we know what ant consciousness is, in order to confirm if we've replicated it or not? It should be possible to get a one-to-one match between basic brain functions, so that we can know if the ant is sleeping, is angry, is hungry, etc., based on what we know of our own neurology. We might not "feel" exactly what it's like to be an ant, but we should eventually at least be able to tell its intentions and brain state on the neurological level.
rosencrantz Posted October 3, 2015 Posted October 3, 2015 One of the fathers of sociobiology (and also a proponent of r/K) E.O. Wilson wrote the definitive book on ants. If you are interested in those bugs, pick it up. The pictures and the layout are quite beautiful. He makes an astute observation, namely that Socialism can work for ants. One of the prerequisites though is the lack of reproductive independence. When you can procreate with a mate, the best evolutionary strategy is to look for your own offspring.
Recommended Posts