I want to make some claims about the future of brain cognition that will lead to rational, logical, empathetic thought. The notion of “friendly SAI” and “unfriendly SAI” is a fallacy and should be abandoned, that is, the notion that we can program AI in the SAI setting to be friendly is an attempt to undermine intelligence and the domain of empathy and altruism.
We must assume however several important properties future brains will have (digital and/or biological).
-
Posthumans (SAI and transhumans) will have superior logic, reason and critical thinking.
-
On top of being super-rational they will also probably have direct access to the internet via brain to computer interfaces, perhaps built in wifi like technology to get information much quicker than we humans do today
-
We should also assume that future brains will be able to multitask past it's current limit of two goals. [1]
-
The future brain will also probably have more memory
-
Brains will also be conscious and aware of both their memory and their current thoughts well beyond the capacity of the human brain.
SAI and posthumans will have two important resources at their disposal – consciousness and the ability to rationally think about theories and concepts. Philosophers have come up with many different forms of ethical and moral theories. Any Being that has consciousness and the ability to understand moral and ethical theories also has the ability to act on them in the real world.
With greater cognition a mind will be able to critically think about their stance on moral issues, and then take action in the real world. Many make the claim that this will not happen because it rarely happens with humans, but we are not talking about humans here. Humans may be hardwired genetically for aggressive behavior. This does not mean we cant take a lesson from the human mind though: humans can be immensely empathetic and altruistic.
Where do these empathetic and altruistic thoughts and actions come from? If social-Darwinism is hardwired in the brain, then is it a fight for us to be friendly? Do we struggle with the notions of being empathetic and altruistic ourselves?
Ayn Rand has been quoted saying many interesting things (most of which lack empathy but do indeed have some kind of “intelligence” to their rationale. For example she said "The ladder of success is best climbed by stepping on the rungs of opportunity." and "Run for your life from any man who tells you that money is evil. That sentence is the leper's bell of an approaching looter." Rand is trying to tell us that we should be out for ourselves, and that Capital, in the Marxian sense is good. Well, the relationship between consciousness and capital is good to understand, like understanding individual success in a social-darwinism driven world. I would argue however that these concepts are building blocks towards altruistic and empathetic behavior.
If one understands how horriable the world would be if Rand's words became true, one would also understand what an empathetic and altruistic world would look like. (I am not endorsing looking into Ayn Rand's words. If you do, read at your own risk.)
There is no doubt in my mind that SAI and the posthuman will see the words of philosophers and great thinkers as stepping stones towards a more “enlightened” wolrd and understanding our place in the universe and the fact that we all live together on this tiny planet of ours. SAI and the posthuman will be empathetic and altruistic not out of some social Darwinian necessity, but because that is where information, intelligence, and critical thinking will lead them. Thank goodness that this type of thinking and behavior may even save us from destroying ourselves.
Perhaps social-Darwinism (in the consciousness domain) is in fact the ability to see humanity as a whole – to critically think ourselves out of being selfish, destructive, and greedy?
References
[1] http://www.scientificamerican.com/article.cfm?id=multitasking-two-tasks
[2] http://www.theatlantic.com/technology/archive/2012/11/noam-chomsky-on-where-artificial-intelligence-went-wrong/261637/
IEET Blog
Interesting article, Kris. I have a couple of questions. First, this article assumes (rather than argues) that SAI will exist. Why? Are you relying on Kurzweil and Moravec’s arguments? Even those seem to me to be problematic because there are lots of environmental, economic, and existential hurdles to our very existence as an advanced civilization. See this article published today, for instance, which references a group of academics: http://news.yahoo.com/blogs/ticket/world-2030-u-declines-food-water-may-scarce-162757458—politics.html It gives good evidence about those hurdles. Also, why see such things as altruism as a process that is not primarily genetic (like aggression). There is plenty of evidence that it might be (ants exhibit it, for instance), because it enhances a species’ survivability. I ask these in the spirit of genuine philosophical debate, and wonder if you could amplify a bit. I’d like to be optimistic about AI and our future, but I’m still on the fence.