Friday, 10 July 2015
SMARTS Update: When Smart Things Go Wrong
A crucial insight struck me when I was researching my new book, SMARTS, which is about the nature of intelligence, who and what displays it, and how we learned to manufacture it in machines.
I attended a summer school organized by a professor of psychology at the Université de Montréal. It was in honor of the centenary of the great Alan Turing who kicked off our smart era by inventing the intellectual principles behind the first electronic computer. (That makes Turing sound very grand, yet his writings display a puckish and earthy sense of humor almost never seen in scholarly humanities journals, let alone those concerned with abstruse subjects like mathematical logic. When Turing wrote an article for Mind about his ideas for an artificial intelligence, he explained that the kind of smart computer he wanted to make -- that could fool us into thinking it`s human-- wouldn't need to be housed in a human-like robot because it wouldn't be asked to fetch coal with the scuttle.)
As I arranged my notebooks and pens on a desk in that lecture hall in Montreal, it seemed to me that there was something odd about the other students and professors. But what? I looked around the room. Hah! Nobody had notebooks or pens. Nobody but me. Everybody else had set out at least one smart machine (an i-Pad, a computer, an i-Phone) to record the meeting. While I was busy getting ready to train my memory of the lecturers' talks by the use of my note-taking hand, they were machine multi-tasking--on their phones, sending texts, emails. In other words, I was the fossil, while they were modern. I would leave the lecture hall hours later with numb, ink-stained fingers and a whole lot of material tucked into my brain for future use. They would leave with the whole event available on a machine whenever they wanted it. Their phones and computers had replaced the need for hands to make notes and for minds to remember. And that`s when I really arrived at the core of my book. I understood that we aren't just inventing and using smart machines, smart machines are forcing us to adapt to them. In a way, they are re-inventing us, body and mind.
While all TV and radio journalists record interviews and events with smart machines these days, and most newspaper people do so too, magazine journalists of my generation rarely even lug small tape recorders around. Our interviews tend to run for hours, unlike those whose end product is a short clip or story. Why record and transcribe hours of interviews loaded up with ums and ahs? We take notes by hand because it's efficient. That`s why most of us also developed quick ears to catch that quotable line as it falls from the lips of hapless subjects, and taught ourselves to get it down on paper, accurately. But active note-taking does more than record words spoken. It also helps one to remember and to process meaning, especially when the material is surprising. It slows the conversation down to the speed of the hand so the subject's answers have a chance to sink in. It gives one time to generate good questions, not just those brought to the interview. Taking notes sets the stage for real conversation. So taking notes helps you learn about someone else`s realities.
And yet, having made those notes, we read them once and put them aside. I learned to do that early on. Why? One day, a senior writer I worked with spotted me in my office, surrounded by open notebooks, struggling to write a story. It was going nowhere good. He read it over my shoulder. I had filled it with long but accurate quotes straight from my notes. Garble, he said. Boooorrrrrrinnnng! Put away your damn notebooks and tell it your way. You're getting trapped by the way people talk which hardly ever is exactly what they mean because almost nobody talks as well as they write. Besides, anything you should use, you already remember. Because it stands out from the crap.
I tried it. He was right.
That moment in Montreal taught me that my old ways and the days of hands teaching minds have gone. They have been replaced by a new dynamic, driven by the ubiquity of smart machines. A dynamic is a two way street. One side reflects its interaction with the other. Both change the other.
There is nothing new about technology changing us. The archaic ancestors of modern humans had much heavier skulls and jaws and much more robust teeth than we do because they processed their food with nothing more than stone and fire. They chewed grit with lunch. They needed big muscles and big muscles need robust bone. The human form changed again about 10,000 years ago with the invention of agriculture and the first settlement of villages. Physical anthropologists can see those changes in the skeletons found in ancient cemeteries. The bones of archaic nomadic people do not show much evidence of settlement/agricultural diseases like tuberculosis, or dental abscess. The human beings who lived off agriculture were physically different from their pre-agricultural cousins. They were changed by new tools invented and the new cultures they permitted. Once again, the shapes of skulls changed.
Our minds changed too. The human capacity to remember was altered by the invention of writing. Nowadays, no one in any western society-- except RADA trained actors and first rank opera singers-- is able to remember long scripts and poems as well as pre-literate Greeks, some of whom could recite Homer's epic poems word for word. We are so distant from that sort of memory capacity that most 20th Century scientists failed to consider peoples' memories as a useful source of information when working among pre-literate societies. Only now are archaeologists, geologists, and especially lawyers in Canada turning to First Nations' oral traditions to answer vital questions. Their stories, agreements, and observations of nature, handed down word for word from one generation to the next, have turned out to be as accurate as any modern, written history. It was an oral history of the Inuit that led scientists to the sunken ships of the doomed Franklin expedition, lost for 170 years beneath the Arctic ice.
There are those who argue that the spread of literacy, thanks to the printing press, changed minds more than the invention of writing. After all, for thousands of years in most western cultures only a few learned to read because books were written and copied by hand. They were so expensive only the rich had libraries. More than one scholar argues that IQs are rising worldwide now due to the global availability of smart machines and the modes of thought necessary to make use of them, not to mention the information they make available.
In other words, human intelligence is being reshaped by intelligent machines. While they remove the need to remember the words we write, and the speeches we give, they force us to remember other things entirely, such as long strings of letters and numbers for our ever shifting passwords. We must also learn and become expert in constantly changing procedures as we interact with ever improving software. The use of my hand to make notes engaged my attention and my memory. The use of smart, searchable machines makes the need for the human memory of words redundant even as it fractures attention constantly. Every day I have to remember how to make my damned machines work, and then I have to shut out their infinite possibilities for entertainment so I can focus on the work I need to do.
They are changing us.
As we are changed as individuals, our communities change too.
This week, our communal dependence on smart systems became obvious for all to see. On Wednesday, disaster struck the New York Stock exchange where billions of trades happen on a daily basis. Its system suddenly crashed. It stayed crashed for over three hours. Then the United Airlines' computer system crashed too. Hundreds of flights were cancelled or delayed. Many thousands of people had their lives disrupted. What were the causes, glitches or terrorists? The head of the FBI worried that terrorist/ hackers were showing what he/she/they could do. But it turned out to be glitches. In a way, glitches are more frightening than planned acts of malice. You can prepare for them and you can sometimes predict them. What you can't predict and plan for are the random glitches that all smart systems are prone to. They will happen no matter how careful we are, no matter how many back up systems we employ. Yet our cities cannot function if the electrical systems go down, and, as we learned in an ice storm some years ago, one system's fault can spread widely to many others. And what is the sine qua non for smart machines? Electricity.
And what did Murphy tell us about all complex systems?
What can go wrong, will.