There is never going to be
artificial intelligence.
by Norrin Radd
How many lines of codes are needed to make these two dots — to the right — chase each other?
If you want, sure, A.I. has already arrived. You could say that computers 'understand' input, that they can 'read' and also 'write', even 'speak' today. Maybe one day the Hallmark of humor will be challenged by an advanced form of pattern recognition, that will enable computers to recognize jokes. And of course they could be programmed to 'laugh' or even choose to not laugh, context-based.
At authentic internet, we're trying hard to make your CPUs look bright. Now and then, we may succeed to make you wonder, how we made the computer do that, without you having to tell it anything about the task at hand, like the smarts of Solon. But in the end, it's always just a good set of rules, a sensible abstraction of the problem, a self-adjusting system. It can get quite helpful (and annoying, too, like some infamous 'assistants'). But the computer never really understands anything.
Why? The more complex systems get and the more layers of abstractions are involved to gain compatibility and power, the less predictable have machines become. Internet technology is a good example for this: many technical layers, high versatility and unpredictable drop outs, lags, crashes. This development towards chaotic behavior is the strongest tilt to showing a human kind of inconsistency that machines have revealed so far. And your PC really always tries his best. The complex systems of today even need immune systems to protect them against parasites. Although the anthropomorphic labeling of 'viruses' and 'worms' makes them appear more tangible than they really are, the underlying pattern and problems are in fact resembling the name-giving biological phenomena.
resources