AlphaGo wins! Has AI arrived?

No! AlphaGo is just a tool designed by humans. Like a vehicle or an industrial machine, it beats the bare-handed human at a certain task.

AlphaGo is just an extension of its designers’ intelligence. Could humans ever hope to beat AlphaGo in a rematch? No doubt. Given enough resources, a competing human team will be able to build a better machine that will beat AlphaGo. And that machine will also be just an extension of human intelligence.

In the recent self-driving accident that killed its occupant, Tesla is wholly responsible. The company has not attempted to blame the accident on the “autonomous” AI. Similarly, Google will be fully responsible if AlphaGo goes awry, just as Microsoft had to retire and fix the “racist Nazi” Tay chatbot.

What about the singularity? Will machines take over the world eventually? Unlikely. Conscience and willpower are not just regular milestones in a linear progression.

We can augment our physical power well beyond our natural limits by applying intelligence, but there is nothing else that we can apply to our intelligence to build it beyond our natural capabilities. Speed, accuracy, endurance and other qualities where computers surpass humans are helpful, but are not intelligence.

Posted in News/Books and tagged , , .