It’s not ‘AI,’ and it’s not ‘superintelligence.’ So, what is it?
Fast Company recently ran this article. Lots of questions. There are “answers” to some of the questions.
Here is an example of a reply from a contributor.
A fair question.
Machines being able to outperform humans on various classes of tasks is nothing new. That’s why we have computers. They originally earned their place with the ability to perform the calculations needed to accurately hit moving targets fast enough to be useful in real time. That’s still important today as we routinely use control systems that require computational performance far beyond human ability.
Computers are tools. They extend human abilities and knowledge. But they are not people. With machine learning they can exhibit behaviors very similar to a limited set of human behaviors. These behaviors benefit from the ability to gather and organize intelligence from a wide variety of sources, in particular benefiting from improved semantic and grammatical data bases. I use the word “intelligence” in its meaning as the collection of, or collected, information, viz. the “Central Intelligence Agency”. Machine learning systems are not autonomous. They require extensive human activity to provide and maintain the environment in which they can function, as well as direct their activity.
We also use “intelligence” to describe the operation of animal brains. We can say with high confidence that animal brains are not Turing machines, not computers. These organic brains are completely autonomous. Their behaviors are not algorithmic, but complex and generally unpredictable.
So we clearly have an ambiguity to resolve. It is clear that machine learning is a toolset which extends and enhances human intelligence. We need a new word. I’m thinking about it. I like the way the Germans have named the smartphone, it is called a “handy”…..
