Artificial Mediocrity, The Next Necessary Thing
This is a reply to a fantastic post named Artificial Stupidity: The Next Big Thing. by Nova Spivack.
In this post the author presents the following ideas:
- He doubts that computers will be able to make intelligent decissions (comparing to humans).
- We should concentrate mainly on low level thinking software (organizing, tagging, linking, remembering, etc.).
- We should let humans making complex thinking, decission making, learning, teaching, problem solving, etc. etc.
The big point in this post lies in reducing our aims. Humans are working on AI since 1950s, and the Nexus 6 has not been built yet. And we are not going to build it soon.
I think most of people working with Semantic Web has this idea in mind. We don’t expect the web to sing Daisy Bell as HAL did, but perhaps humans are not so intelligent as Nova states, and computers could be smarter than he says.
I will follow the knowledge stairs. In the first floor we have raw data, which can be represented by tags and populations. Humans have a nice, but limited data store. Computers can memorize huge amount of data.
Computers 1 - Humans 0
When associations are made between data, information is obtained. So information links data in some way. Computers are great (and improving) at dealing with information (storing, linking, retrieving, etc.)
Computers 2 - Humans 0
In the next step we get Knowledge from information collections organized in a useful manner. A rule is a kind of knowledge. Computers can deal with some kind of rules, but it’s being really difficult to develop these systems (Semantic Web):
Computers 2 - Humans 1
And the last step is understanding, an analytic and probabilistic process that allows us to learn knowledge and syntetize new knowledge. Just a Nexus 6 can do it, so:
Computers 2 - Humans 2
It’s difficult to define the concept intelligence. It is easier to talk about intelligent habilities, as learning, organizing and processing data, and aplying learnt knowledge to resolve complex problems.
Computers already have some intelligent habilities. And they get more each day. And thanks to them humans (not necessary very intelligent) are able to solve more complex problems. Computers are getting their information age maturity, and knowledge age is beginning now.
In conclusion, we have artificial stupidiy yet. We should go on with Artificial Mediocrity.
