AI point-missing at a somewhat higher level
From MindMatters as usual.

Michael Egnor: Sure. I mean, I’ve always thought of artificial intelligence as just representation of human intelligence. And, in a sense, that the term artificial intelligence is an oxymoron. If it’s artificial, it’s not intelligence.

Robert J. Marks: What you mention is exactly the tests that Selmer Bringsjord, a professor at Rensselaer used in his test for whether artificial intelligence would be creative. His test was, does the computer program do something, which is outside the explanation, or the intent of the programmer? And there has been, thus far, no artificial intelligence that has done this.

This set of definitions skips an important step.

Humans can make two very distinct kinds of tools, and use those tools in distinct ways. When we use any tool, or when we perform any action without a tool, we're capable of creativity. Most of the time we don't create, we just do repetitive stuff that gets us through life.

The discussion misses the distinction between kinds of tools.

We've always made and used simple tools, in the same way that birds or monkeys can make and use simple tools. Pick up a stick, pull off the leaves, use it to poke into a hole and dig out bugs, or use it to draw a picture in the dust.

In the last few thousand years we've started to make tools that have their own intelligence. A clock or a mousetrap or a computer is a tool infused with intelligence by its human maker.

Unlike a stick or a pencil or a typewriter, clocks and mousetraps and computers perform tasks when we're not physically pushing them. The clock ticks regularly and "knows" when to strike 12. The mousetrap waits patiently and "knows" when to snap and kill a mouse. The computer ticks regularly and "knows" when to run a subroutine. There's nothing special about a computer, even when it's running an AI program. Clocks have been running complex programs with video and audio displays for 800 years.

All of these infused tools can be creative when something goes wrong. Clocks can strike 13. Mousetraps can respond to wind gusts. Computers can reach wrong conclusions.

Human creativity is the same. It usually happens in the middle of a repetitive tick-tock task. You're washing dishes or adding a column of numbers or walking, and suddenly an internal gear slips and makes a connection between two unrelated subjects.

The difference between the clock striking 13 and the unexpected connection isn't in the unexpected action itself.

What makes us ALIVE is our ability to PERCEIVE that the unexpected connection is USEFUL. The clock doesn't stop ticking and find a way to use the 13th ding. We do stop washing and write down the unexpected idea, then we start a new repetitive process to develop the unexpected idea into a new habit or a new painting or a new business.

Effective creators have well-formed repetitive habits, well-worn paths equipped with their own tools, for developing those connections in our habitual direction.

LIFE IS PURPOSE.

= = = = =

Metaquestion: How does a repetitive task stir creativity? Possibly because repetitive tasks like walking create new neurons, which immediately start to perform their intended PURPOSE of seeking connections. Because the new neuron isn't already linked into existing connections, it WILL form new connections.

Metametaquestion: Why did God want us to create while rhythmically repeating habitual tasks? Probably unanswerable, but might be worth some thought.

= = = = =

Later and bigger question from a different angle. Starting with the printing press, automation has been stripping away REPETITIVE PHYSICAL TASKS. That's the main selling point of automation. 200 years ago we washed clothes and swept floors with rhythm, we walked more, we wrote with our hands more, assembly-line workers and bookkeepers performed more repetition. Every new "labor-saving" device eliminates more of the physical rhythm and repetition that gives us new neurons. Is this why we haven't made any REAL inventions in the last hundred years?

The question was inspired by reading this sales pitch for automation in a 1916 book about adding machines:

"Curious, isn't it, when you stop to think about it. Man can make mistakes and does. Yet man, who makes mistakes, makes the Bookkeeping Machine which does not make mistakes. And with the machine there is no brain blow out, no wear and tear on the cerebellum."

Exactly backwards. Using the cerebellum expands the cerebellum. The machine can make mistakes. When the machine makes mistakes it doesn't lead to improved life. When the human makes new connections it can lead to improved life. But if we aren't PERFORMING the repetitive task of bookkeeping, we DO wear out the cerebellum and we LOSE the opportunity to create. The brain's glial janitor service is always at work, always sweeping out neurons that aren't participating in connections.

Labels: AI point-missing

  ¶ 2:17 PM