Galileo and the Dolphins

by Adrian Berry

  Driverless cars, although in their infancy, promise a vast reduction in the number of accidents within a few decades. Since almost all accidents are caused by drivers, the ideal solution is to abolish them.

  There is a precedent for such optimism in the chess-playing machines that routinely defeat all but a handful of grandmasters. Road conditions, for all their possible variations, are considerably less complicated than positions on a chessboard. A chess machine is programmed to remember a huge number of positions, knowing from past experience how to react to them. It seems overwhelmingly probable that the hitherto baffling problems of road safety can be solved in the same way.

  An electronic driver, like a chess machine, would remember past ‘positions’. Confronted by a patch of oil on the road, a drunk lurching in front of the car or a bouncing ball indicating the presence of playing children, it would know how to react.

  ‘But cars that drive themselves so safely that you can tell them your desired destination and go to sleep are still at least ten years away,’ says Pomerleau. ‘One great electronic improvement which should prevent a lot of accidents could, however, be ready in as little as three years. This is a warning system that wakes drivers up when they start to fall asleep.’

  About half the motorway deaths in the Western world are caused by drivers falling asleep at the wheel. It is a problem as serious as drunken driving. A gadget operated by electrodes planted in the driver’s skull that would beep if he showed signs of drowsiness would be a tremendous step towards death-free roads.

  In a short story half-jokingly suggested that a major problem of the future might be the wild social life of intelligent cars. They would go out by themselves to parties where they would ‘mate’ and produce young. While this may never happen, we can visualize situations in which they are independent beings as seemingly intelligent as chauffeurs and immeasurably faster in making decisions.

  I’ll Be Back!’

  Robots like the Terminator, programmed to hunt and kill particular people, could be feasible within a few years. They may not look quite like Arnold Schwarzenegger in the cult films but they could be just as deadly.

  ‘Already it is technically possible to build a robot that would roam about and kill everyone it met who had blonde hair,’ said Kevin Warwick, head of Reading University’s cybernetics department, who lectured to a packed hall on the darker side of machine ‘intelligence’ of the future.

  ‘It has been argued by some scientists, notably Roger Penrose in his 1989 book The Emperor’s New Mind, that such machines are theoretically impossible because they could not be made aware of their own existence.

  ‘But I believe that view is wrong and that such books are less than responsible. To act with intelligence, a machine does not need self-awareness. It only needs a purpose,’ says . ‘A good example of a machine with such a purpose is a chess-playing computer. It can even defeat its own creator. It has the overriding goal to win the game, yet it does not even know it is playing a game.’

  and his colleagues have built a team of miniature robots with insect-like intelligence which he calls the Seven Dwarfs. Each has a computer chip for a ‘brain’ and ‘sees’ with ultrasonic eyes. They can be programmed either to hunt each other or to flee from each other across his laboratory floor.

  ‘In a few years’ time, to judge by present advances in the speed and memory of computers, it will be possible to build a much larger version of such a robot, for police or military purposes, but which a criminal could programme to seek and assassinate his victims.

  ‘It is essential to understand that computers and robots do not think like we do and, as Penrose has correctly pointed out, probably could not even be programmed to do so. They do not have any in-built capacity for humanity or mercy, and it is idle to think that we can protect ourselves from them by any device like Isaac Asimov’s proposed first law of robotics , which states that a machine must not harm a human being.

  ‘We set the machine a goal, expecting it to execute the goal in our way. But instead, it executes it in its way. A superb example of this was seen in the 1969 film The Forbin Project, in which a computer was instructed to take over the world’s nuclear weapons with instructions to bring about peace . But it achieved peace by massacring nearly everyone.’

  Meanwhile, experimenters at the Information Research Laboratories in Kyoto, Japan, are trying to build a robot with the brains of a cat. This does not mean that it will be furry or chase mice or want saucers of milk, but that it will do things without being told how to do them. It will program itself and then execute its own program. It will be the first robot that is more intelligent than an insect. Yet, devoid of morals or common sense, such robotic ‘animals’ are all too easily liable to behave in ways that are totally at variance with what their builders intended.

  believes that too many people do not appreciate the danger of such misunderstandings. They rely too much on the supposed wisdom of computers, he says. Banks, for example, have machines that decide whether a particular person should be given a loan. But the computer is liable to reason in ways in which its designer never intended.

  ‘For example, it might have twice before rejected loan applications by people who happened to be left-handed. It could then decide, without announcing that it was doing so, to reject all applications from left-handers.’ It may be a small step from stupidity to crime.

  Reaper and Creeper

  Computer scientists from Hebrew University in Jerusalem flew home from a recent conference in Cincinnati to some very unpleasant news. They had been telling their American colleagues of their fears of electronic viruses - malicious software inserted into computers that wipes out files and programs.

  On their return they found that such a virus had attacked all the interlinked personal computers in Jerusalem. Fortunately the villain had made two errors - one technical, the other religious.

  He or she had left a tell-tale trace of the attempted sabotage, so that when any of the affected computers were switched on, an abnormal message flashed briefly on the screens, indicating that there had been some tampering with the programs.

  The virus was due to activate on Saturday, Israel’s Sabbath, a day when few people would be working at their terminals. There was plenty of time to warn them through the newspapers. A plot that might have done significant damage to Israel’s economy had been foiled by the incompetence of its perpetrator.

  Computer viruses are quite different from hacking - the malicious art of computer snooping. The viruses are maliciously designed to do the maximum damage to a computer system by destroying all the software.

  In many ways they work exactly like their biological counterparts. A virus may attach itself to a program or a file in such a way (unlike the effort of the Jerusalem saboteur) that it cannot be detected. It can then make a copy of itself which it transfers to the next piece of software before destroying its original host. It continues this process until everything in the computer’s memory disks has been wiped out.

  Another kind of virus does the opposite. Instead of destroying everything in memory, or in the memory storage disks, it gradually fills up all the free space with garbage, so that eventually the computer cannot be used.

  In one example a programmer feeling aggrieved after being sacked from a large company left a ‘gift’ behind him in the company’s computer. It was a tiny program called Creeper, occupying no more than 400 bytes, that had only one function: to make an exact copy of itself.

  Twenty-four hours after the sacked programmer had left, it woke up, duplicated itself and went to sleep again. No harm done yet. A massive computer with a memory of 300 million bytes had merely been invaded by a paltry 800 alien bytes of coding. But two days later the two copies of Creeper again procreated. Now there were four of them, totalling 1,600 bytes. So the process went on, day after day ... 8, 16, 32 copies of Creeper, remorselessly eating up memory space. And still nobody noticed they were there.

  At the end of a fortnight strange things
started to happen to some of the peripheral functions of the company’s computer. There were odd delays and apparent mistakes. For its memory now contained 16,384 identical copies of Creeper -more than 6.5 million bytes of utter rubbish.

  After 20 days the computer refused to perform any task whatsoever. Its memory was overwhelmed by the presence of more than half a million copies of Creeper.

  In desperation the company created emergency memory space into which they introduced a program of their own called Reaper. The task of Reaper was to hunt down all copies of Creeper and destroy them.

  The war between Reaper and Creeper was like that between Hercules and the Hydra. No sooner had copies of Creeper been destroyed than more appeared. The battle between the two warring programs - in which Reaper was eventually victorious - prevented the computer from carrying out its normal tasks and very nearly bankrupted the company.

  But at least the struggle has led to the creation of a new kind of computer game, Core Wars, in which two opposing programs struggle to destroy each other. They even fire at each other with a form of electronic artillery.

  Electronic ‘vaccines’ have been invented to combat viruses. One such product is Safeguard, being sold by the London firm Prosoft. ‘Safeguard will check a legitimate program to make sure that no alterations have been made to it since it was last run,’ says its manager . But even Safeguard won’t find new programs that have been maliciously inserted. Only constant vigilance and a suspicious mind will do that.

  It is unlikely, of course, that a virus could dwell in legitimately sold software. But too many floppy disks are circulated privately, their recipients not bothering to inquire about their origins. Therein lies the danger. That is why experts liken computer viruses to Aids.

  A Cable Car to the Heavens

  Sharp-eyed people in tropical countries were amazed one week in 1994 to see a long, thin line extending across the night sky, shining faintly in reflected sunlight.

  This was no UFO but an orbiting man-made ‘tether’, 20 kilometres long and only an eighth of a centimetre thick. It was made in the garage of Joe Carroll, founder of a small Californian group of space enthusiasts called Tether Applications, and launched into space by hitching a ride on an unmanned Delta rocket, a feat that astronauts in space shuttles had been attempting for nearly a year with much less success.

  ’s tether consists of a form of polyethylene, the tough plastic of supermarket shopping bags. Scientists are watching his experiment, which could have profound consequences for industry in the twenty-first century.

  For many years there has been a growing realization that space technology will be cheaper and more efficient if the work of rockets is supplemented by other devices. A space tether could perform a vast number of tasks, from providing electricity to spacecraft and throwing cargo between the Earth and the Moon, to ensuring the good health of people on their way to Mars.

  In its simplest form, a tether could be a mini power station in space. A tether made of semi-conducting material with one end passing through the Earth’s magnetic field could carry electricity to a spacecraft, supplementing the work of solar panels which normally power such a craft by direct sunlight.*

  *ln 1996, shuttle astronauts trailed a 20-kilometre tether in the Earth’s ionosphere, trying to extract electricity from it. But to everyone’s great disappointment the tether snapped, apparently overloaded by the electricity that was moving along it.

  ‘A tether could also heighten or lower the orbit of a spacecraft,’ said . ‘Extending from a spacecraft in low Earth orbit, it always wants to point downwards, towards the atmosphere. When its downward end enters the atmosphere, it will be affected by drag, slowing the craft and pulling it downwards.

  ‘Conversely, if the same tether is suddenly cut, pressure will be released and the spacecraft will bound upwards.’

  A more ambitious scheme has been proposed by , former chief scientist at the Hughes Research Laboratories, who sees tethers as a future Earth-Moon transport system. ‘We hope to hurl cargo weighing up to five tons nearly half a million kilometres across space, rather in the way that David used a slingshot to kill Goliath.’

  Forward’s idea goes like this. Cargo from a space station circling the Earth is to be sent to the Moon without using rockets. The Earth station has a tether extending to the cargo at its end. When the cargo faces the Moon, the tether releases it. The cargo is then caught by another tether attached to a space station orbiting the Moon - like a baseball player catching a ball - and lowered to the lunar surface. ‘This system would be ideal for sending to the Moon essential supplies like scarce hydrogen, with which lunar colonists could manufacture water.’

  People travelling to Mars would risk ill-health because of the effects of weightlessness. The progressive loss of calcium in their bones can accelerate ageing, and would make them unfit for work on reaching the Red Planet.

  The solution would be a rigid tether a few kilometres long. One end would be attached to the Mars ship and the other to a counterbalancing object weighing roughly the same as the ship. By the firing of small rockets, the whole system, ship, tether and counterbalance, would be made to rotate. As on a fairground ride, centrifugal force would create gravity artificially inside the ship, on the side opposite to the tether. By varying rotation speed, gravity of any desired strength could be created on board. A problem would only arise if the ship had to do a mid-course correction.

  The most remarkable suggestion of all for a space tether comes from , who describes what would be a cable car to the heavens. One end of a 36,000-kilometre tether would be attached to a point near the equator and the other to a satellite in geostationary orbit.

  A lift would then travel up and down this taut cable, carrying people and freight into high orbit. (The problems and perils of building such a ‘cable car to the stars’ are graphically described in his novel, The Fountains of Paradise.)

  A ridiculous notion? ‘I predict that it will be built about fifty years after everyone stops laughing,’ says .

  And the 1994 experiment? said: ‘My tether was severed in one place by a micrometeorite, but not until it had been in orbit for five days. I count this a success, with this proviso: people using space tethers will always have to allow for occasional breakages.’

  The Devil’s Digits

  Mighty are numbers, joined with art resistless.

  There may be dangers ahead for those who have put their faith in the mightiness of numbers. A secret code whose security depended on sheer numerical complexity has been found wanting. It appears that once several hundred computers get to work simultaneously, then even the most cunningly hidden mathematical secrets will be revealed.

  The widely used RSA cipher (named after its inventors Rivest, Shamir and Adelman), works on the assumption that if two very large prime numbers are multiplied together, it will take centuries - millennia even - for the fastest computer to do the sum backwards and identify the two original primes which conceal the secret keys to the cipher.

  Take the ‘Evil Number’ 3,458,137, a discovery of my own. It takes a home computer less than a second to discover why it is ‘evil’. After several hundred calculations performed at lightning speed, we find that it is the product of the primes 1789 and 1933, the respective dates of the onset of the French Revolution and of Hitler’s rise to power - both thoroughly evil dates.

  But if the two prime numbers (divisible only by themselves and one), are not of four digits but more than 100, then, so the three mathematicians reasoned, the problem would be truly ‘intractable’. The would-be codebreakers would long have died of old age before they found the hidden figures.

  However what they did not reckon with was that ‘prime factoring’ - as it is called - would become a cult. Looking for hidden primes has become one of the standard tests of a computer’s speed. In 1994, a team of mathematicians, using a network of more than 400 computers around the world, succeeded in factoring a number of 129 digits. This spectacular number, known appropriately as
RSA 129, was:

  114,381,625,757,888, 867, 669, 235, 779, 976, 146,612,010,218,296, 721,242,362,562,561, 842, 935, 706, 935, 245, 733,897,830,597, 123, 563, 958, 705, 058, 989, 075, 147,599,290,026, 879, 543, 541

  It was divisible, the ‘crackers’ announced as their champagne corks popped, by the two primes:

  3,490,529,510,847, 650,949, 147,849,619, 903,898, 133,417,764, 638, 493, 387, 843, 990, 820,577

  and:

  32,769, 132,993,266, 709,549,961,988, 190, 834,461,413, 177,642, 967, 992, 942, 539,798, 288, 533.

  It was a feat that could have taken a single super-computer as much as twenty years.*

  *And, so after all this, how safe is the RSA cipher? The answer is that it is almost certainly safe - for the present - if a prime product of at least 150 digits is used. But this is only true if one is sending messages and files that do not have to remain secret for very long. It is not safe for archival files, that must preserve their secrecy for many years. Advances in factoring will continue to be made and ever larger numbers will become easy to factor quickly. And yet the whole RSA cipher could be destroyed by the invention of ‘quantum computers’, when that invention comes.

  started the whole thing more than 2,000 years ago with his somewhat obscure proof that there is an infinite number of primes. ‘Every whole number greater than one,’ he then announced portentously, ‘is either a prime or it is not. And that shall be called the First Law of Arithmetic’ To which a critic sarcastically replied: ‘Every person aged greater than zero is either male or female. And that shall be called the First Law of Humanity.’

  Eratosthenes of Cyrene worked out an ingenious ‘sieve’ to discover all prime numbers up to a given limit. As the rhyme goes:

  Strike the twos and strike the threes,