Written by Quentin Langley
Let us suppose that a computer programmer with an IQ of 140 can devise an artificial intelligence with an IQ of 180. Let us further suppose that this genius computer is given the task of programming a new artificial intelligence with a still greater IQ. Might it devise an AI with an IQ of 230? And what if this computer were given the same task?
This is the speculation that led to the concept known as the Singularity. Each new generation of AI will be smarter than the previous one, and though it has taken our species 10,000 years to get to the point of the first generation, the computers might be able to create new generations in a few minutes. Suddenly, the maximum intelligence on the planet can multiply at a fantastic, and possibly alarming, rate. We will have reached a point where technology can advance unbelievably fast.
People are not comfortable with rapid technological and social change. The writer, Douglas Adams, defined three types of technology: 1. Tech invented before you were born, which is not really technology at all, just part of the natural order of things; 2. Tech invented between your birth and your 35th birthday, which is cool and wacky, and you might be able to make a career out of it; 3. Tech invented after you turn 35 which is dangerous and wrong and needs to be banned.
We are all familiar with this view. Parents who blame the internet for damaging children were, in turn, damaged (in the view of their own parents) by TV. And those parents were damaged by radio. It was probably the case when our species conquered fire.
Is this why people fear the Singularity? Stephen Hawking and Elon Musk are said to be among the doubters. When the greatest intelligence on the planet is no longer human, but machine, will our machine overlords begin the hunt for Sarah Connor?
It is remarkable to think that all the technological and social change of recent decades could be dwarfed in the very near future: that humanity – or the machines it creates – could learn more in a few minutes than we have learned in the past 10,000 years: the whole of our history, since the founding of the first settlements.
How can we prepare for a change which is likely to arise in the lifetimes of children now at school? What can we teach children about preparing for a world where machines are going to be smarter than they are?
We can teach them critical reasoning. The Twentieth Century – the age of mass media – was incredible at generating knowledge: more than in all prior centuries combined. But it undermined our ability to engage in critical reasoning. We developed an astonishing superstition: that any data which is widespread or expensive to produce is somehow objectively true. If something is in newspapers, on radio, or TV it is the truth.
On April 1st every year we consume the media critically. We ask, “is this credible?” Are there corroborating reports? Where’s the evidence? We need to do this all year.