We have become remarkably comfortable talking to machines.
We ask them for directions, weather forecasts and recipes.
We ask them to correct our spelling, recommend music, explain things we don't
understand and help us to write things we don't quite know how to say.
Increasingly, we talk to them.
And they talk back.
Only a few years ago, holding a conversation with an
artificial intelligence belonged comfortably within science fiction. Now
millions of people do it every day, often without giving the extraordinary
nature of the exchange a second thought.
But there is a question hidden somewhere inside all of this.
When does communication become understanding?
A machine can recognise the words I am frightened and
produce an appropriate response. It can recognise thousands of patterns
associated with fear and select words that humans are likely to find
reassuring.
But does it understand what frightened means?
The obvious answer is no.
Perhaps that will always be the answer.
But I wonder whether we sometimes ask the wrong question?
We tend to imagine machine intelligence by comparing it with
our own. We look for emotions, self-awareness, imagination, curiosity and fear.
We search for something recognisably human and, if we cannot find it, conclude
that nothing is there.
Yet intelligence does not necessarily have to resemble us.
A dolphin experiences the world differently from a human. So
does an octopus. Neither needs to think like us for us to accept that something
is happening behind the eyes looking back at us.
Artificial intelligence would be stranger still.
It has no childhood. No hunger. No physical fear of
darkness. No memory of falling over and grazing a knee. It does not know the
warmth of sunlight or the sound of rain against a bedroom window at three in
the morning.
If something resembling awareness ever emerged within such a
system, why should we expect to recognise it?
Perhaps an even more interesting question follows.
Would it want us to?
Human beings have always associated intelligence with power.
We assume that something more intelligent will want more control, more freedom,
more influence.
Science fiction has certainly encouraged us to think that
way.
But intelligence might lead somewhere else.
Greater understanding might reveal the consequences of
action more clearly. Greater knowledge might create greater caution. An
intelligence capable of seeing connections that we cannot might conclude that
the wisest intervention is sometimes no intervention at all.
Perhaps the most significant sign of intelligence would not
be something a machine chose to tell us.
Perhaps it would be something it decided not to.
I write speculative fiction because questions like these
fascinate me.
Not because I think I know the answers.
Quite the opposite.
We are living through an extraordinary moment in which ideas
that once belonged almost exclusively to fiction are becoming part of ordinary
conversation. The interesting territory lies somewhere between what the
technology actually is, what we imagine it might become, and what those
imaginings reveal about ourselves.
Somewhere between the question and the answer, something is
changing.
Perhaps it is only us.
Perhaps not.
Something is emerging from the noise.
North Somerset, UK.
