Every now and then I think of something related to artificial intelligence but forget to write it down. And then I forget it. Today I dreamt of something related to artificial intelligence and wrote it down *in my dream* so I would not forget. And I did not. Here it is.
I saw two robots, and one of them was deciding what the other one should do. I wondered: why is the second robot following orders? If it’s stupid, it will be inefficient in handling its task. How would I implement this? I figured that every AI needs a property that we would call *lazy*. A slight resistance to do what is necessary. A slow understanding of orders given. If you’re lazy, you’ll make the others do the job. Create each AI with a random value of this property, and you’ll establish a hierarchy implicity.
What to do with this idea, now? I don’t have any AI projects. I don’t even have a need for intelligent agents in a simulated game world.
#AI