Last week on This American Life there was a short segment by the guys over at Radio Lab about morality. The question was if you could save the life of four people by killing one person would you do it? In the first scenario where by pulling a lever which directed a train you saved the four people, but a fifth died instead nine out of ten people said that they would pull the lever. In the second scenario where to save the four people you pushed the fifth in front of the train the response was overwhelmingly NOT to push the person.

But why? It’s the same question - sacrifice one, save four. But overwhelmingly the response is that killing a person is better/easier/more justified to save others when you don’t have to been the one doing it. To examine why we have such a natural aversion to being the pusher the team at Radio Lab spoke to a scientist who was looking at the biology of moral decision making. Examining brain scans of people making ethical decisions one would expect that there is unique center of the brain that becomes active when making these decisions. What the scientist found was that there are numerous areas of activity competing for the right choice to make.

Upon hearing this I thought of another article I had just read over at National Geographic about swarm thought. Essentially the apparent intelligence of an ant or a bee is the sum of the intelligence of a the ant hill or hive. The result of many, many simple interactions build a complex system of results that suggest a greater intelligence.

An example given in the article is bees searching for a new hive. Scout bees are sent out and when one discovers a suitable location for a new hive it will return to the swarm and perform a dance advocating this new location. (The dance conveys information about where the location of the new hive is). When enough bees have returned and are doing the same dance the location of the new hive is determined. An almost democratic decision made by the entire hive.

Swarm thought and ethical decisions, what do they have in common? They are simple elements that combine to resolve a complex question. And perhaps a more important question, what does that mean about intelligence, artificial intelligence and computing? I would argue that our intelligence is the same as that of a hive of bees or a school of herring. Our understanding of the world is an amalgamation of competing inputs that drive our thoughts and behaviors. The system of competing inputs is the essence of our intelligence and perhaps can be modeled and reproduced.

If intelligence is a system of resolving multiple inputs then we have already created AI, but how many inputs are required for advance intelligence? How sophisticated must the resolution be? I think the future of computing will answer these questions. As our hardware increases in speed and capacity and our software becomes more capable at modeling our natural world we will see if there is anything more to intelligence than marching of ants and the buzzing of bees.