The robots are coming - or are they?
What is the current state of play in Artificial Intelligence?
A. Can robots advance so far that they become the ultimate threat to our existence? Some
scientists say no, and dismiss the very idea of Artificial Intelligence. The human brain, they
argue, is the most complicated system ever created, and any machine designed to
reproduce human thought is bound to fail. Physicist Roger Penrose of Oxford University and
others believe that machines are physically incapable of human thought. Colin McGinn of
Rutgers University backs this up when he says that Artificial Intelligence ‘is like sheep trying
to do complicated psychoanalysis. They just don’t have the conceptual equipment they
need in their limited brains.’
B. Artificial Intelligence, or AI, is different from most technologies in that scientists still
understand very little about how intelligence works. Physicists have a good understanding
of Newtonian mechanics and the quantum theory of atoms and molecules, whereas the
basic laws of intelligence remain a mystery. But a sizeable number of mathematicians and
computer scientists, who are specialists in the area, are optimistic about the possibilities. To
them it is only a matter of time before a thinking machine walks out of the laboratory. Over
the years, various problems have impeded all efforts to create robots. To attack these
difficulties, researchers tried to use the ‘top-down approach’, using a computer in an attempt
to program all the essential rules onto a single disc. By inserting this into a machine, it
would then become self-aware and attain a human-like intelligence.
C. In the 1950s and 1960s great progress was made, but the shortcomings of these
prototype robots soon became clear. They were huge and took hours to navigate across a
room. Meanwhile, a fruit fly, with a brain containing only a fraction of the computing power,
can effortlessly navigate in three dimensions. Our brains, like the fruit fly’s, unconsciously
recognise what we see by performing countless calculations. This unconscious awareness
of patterns is exactly what computers are missing. The second problem is robots’ lack of
common sense. Humans know that water is wet and that mothers are older than their
daughters. But there is no mathematics that can express these truths. Children learn the
intuitive laws of biology and physics by interacting with the real world. Robots know only
what has been programmed into them.
D. Because of the limitations of the top-down approach to Artificial Intelligence, attempts
have been made to use a ‘bottom-up’ approach instead - that is, to try to imitate evolution
and the way a baby learns. Rodney Brooks was the director of MIT’s Artificial Intelligence
laboratory, famous for its lumbering ‘top-down’ walking robots. He changed the course of
research when he explored the unorthodox idea of tiny ‘insectoid’ robots that learned to
walk by bumping into things instead of computing mathematically the precise position of
their feet. Today many of the descendants of Brooks’ insectoid boots are on Mars gathering
data for NASA (the National Aeronautics and Space Administration), running across the
dusty landscape of the planet. For all their successes, in mimicking the behaviour of insects,
however, robots using neural networks have performed miserably when their programmers
have tried to duplicate in them the behaviour of higher organisms such as mammals. MIT’s
Marvin Minsky summarises the problems of AI: ‘The history of AI is sort of funny because
the first real accomplishment were beautiful things, like a machine that could do well in a
maths course. But then we started to try to make machines that could answer questions
about simple children’s stories. There’s no machine today that can do that.’
E. There are people who believe that eventually there will be a combination between the
top-down and bottom-up, which may provide the key to Artificial Intelligence. As adults, we
blend the two approaches. It has been suggested that our emotions represent the quality
that most distinguishes us as human, that it is impossible for machines ever to have
emotions. Computer expert Hans Moravec thinks that in the future robots will be
What is the current state of play in Artificial Intelligence?
A. Can robots advance so far that they become the ultimate threat to our existence? Some
scientists say no, and dismiss the very idea of Artificial Intelligence. The human brain, they
argue, is the most complicated system ever created, and any machine designed to
reproduce human thought is bound to fail. Physicist Roger Penrose of Oxford University and
others believe that machines are physically incapable of human thought. Colin McGinn of
Rutgers University backs this up when he says that Artificial Intelligence ‘is like sheep trying
to do complicated psychoanalysis. They just don’t have the conceptual equipment they
need in their limited brains.’
B. Artificial Intelligence, or AI, is different from most technologies in that scientists still
understand very little about how intelligence works. Physicists have a good understanding
of Newtonian mechanics and the quantum theory of atoms and molecules, whereas the
basic laws of intelligence remain a mystery. But a sizeable number of mathematicians and
computer scientists, who are specialists in the area, are optimistic about the possibilities. To
them it is only a matter of time before a thinking machine walks out of the laboratory. Over
the years, various problems have impeded all efforts to create robots. To attack these
difficulties, researchers tried to use the ‘top-down approach’, using a computer in an attempt
to program all the essential rules onto a single disc. By inserting this into a machine, it
would then become self-aware and attain a human-like intelligence.
C. In the 1950s and 1960s great progress was made, but the shortcomings of these
prototype robots soon became clear. They were huge and took hours to navigate across a
room. Meanwhile, a fruit fly, with a brain containing only a fraction of the computing power,
can effortlessly navigate in three dimensions. Our brains, like the fruit fly’s, unconsciously
recognise what we see by performing countless calculations. This unconscious awareness
of patterns is exactly what computers are missing. The second problem is robots’ lack of
common sense. Humans know that water is wet and that mothers are older than their
daughters. But there is no mathematics that can express these truths. Children learn the
intuitive laws of biology and physics by interacting with the real world. Robots know only
what has been programmed into them.
D. Because of the limitations of the top-down approach to Artificial Intelligence, attempts
have been made to use a ‘bottom-up’ approach instead - that is, to try to imitate evolution
and the way a baby learns. Rodney Brooks was the director of MIT’s Artificial Intelligence
laboratory, famous for its lumbering ‘top-down’ walking robots. He changed the course of
research when he explored the unorthodox idea of tiny ‘insectoid’ robots that learned to
walk by bumping into things instead of computing mathematically the precise position of
their feet. Today many of the descendants of Brooks’ insectoid boots are on Mars gathering
data for NASA (the National Aeronautics and Space Administration), running across the
dusty landscape of the planet. For all their successes, in mimicking the behaviour of insects,
however, robots using neural networks have performed miserably when their programmers
have tried to duplicate in them the behaviour of higher organisms such as mammals. MIT’s
Marvin Minsky summarises the problems of AI: ‘The history of AI is sort of funny because
the first real accomplishment were beautiful things, like a machine that could do well in a
maths course. But then we started to try to make machines that could answer questions
about simple children’s stories. There’s no machine today that can do that.’
E. There are people who believe that eventually there will be a combination between the
top-down and bottom-up, which may provide the key to Artificial Intelligence. As adults, we
blend the two approaches. It has been suggested that our emotions represent the quality
that most distinguishes us as human, that it is impossible for machines ever to have
emotions. Computer expert Hans Moravec thinks that in the future robots will be
