SilasBarta comments on Pain - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (195)
I have a counter-example that seems to show that the badness of pain is not necessarily derived from its effect on decision-making. Suppose I tell you that I'm going to make a copy of your brain and keep it in a jar unconnected to any motor nerves/circuits. I'm going to do this no matter what, but unless you pay me X dollars, I'm also going to keep it in excruciating pain. (If you do pay up, I'll give it a neutral experience.) Assuming you're willing to pay more than 0 dollars, doesn't that show that even if pain has no bad effects on decision-making, it would still be bad?
The example doesn't show "pain that has no bad effects on decision-making". That excruciating pain you're going to put the brain through will fundamentally alter what decisions that brain will make, via diversion of its cognitive resources.
Perhaps you don't consider that effect on that brain's decision making to be bad. Why wouldn't you? If you're using the brain simulation to do human-level AI, you want it at its best. If you're researching how people respond to pain, then it can be good or bad depending on what the purpose of that research is.
But I said that it's not connected to any motor nerves, so it has no decisions to make. Well, I suppose it can still decide what to think about, so let's say I also remove that ability and force its thoughts to drift randomly.
Neither are most supercomputers connected to motor nerves, yet people are quite interested in their output!
In that case, it's no longer clear if "pain" is even well-defined. Something whose thoughts truly drift randomly and are prevented from attaining any order ... is in a state of entropy; it does not even count as a control system, let alone life.