Robots at your service (1981)

From the August 1981 issue of the Socialist Standard

  We are entering into a second industrial revolution, the era of automation. This time it is not man’s muscles that are to be replaced and extended but rather man's brains 

Though the above statement appeared in a book published in 1952, the underlying sentiment is not altogether modern. The chilling thought of man the creator threatened by his own creation is echoed for example in Mary Shelley’s novel Frankenstein, published in 1817.

But the early nineteenth century was the era of the Luddite riots when workers protested not against automation—the word was first coined in 1946 by Del Harder, an executive with the Ford Motor company—but mechanisation. In the 1950s however, when a survey was carried out in Detroit asking people whom or what they feared most, automation came second in the list behind ‘‘the Russians.” But what exactly is automation and what is the nature of the threat it is supposed to represent? In an article in the New Scientist (12/2/81) Peter Marsh explained.

  Every activity has three essential elements: power, action and control. When control is exercised by a mechanical device it is called automatic or self-acting.
  An example is the pressure cooker (invented in 1680). But automatic is not the same as automated. For something to be described as automated requires that it has at least one of three additional features. It must firstly make use of a “systems approach” whereby things are made by passing them through successive stages of a manufacturing process without human intervention. Secondly, the device must be programmable, that is, able to do more than one kind of job. And finally, it must have the capacity to receive and process feedback information through sensors to adjust its routine according to changes that occur around it.

On the basis of this analysis Marsh defines three broad technological revolutions that have taken place over some two and a half centuries (roughly the time capitalism has been in existence).

The first of these was the Machine Revolution (1750-1860). It brought about big advances in both the “action” and “power” elements of activities such as textile manufacture. Next came the Scientific Revolution (1880-1920) characterised by the growth of science-based industries such as chemicals and electrical goods. It was followed after the second world war by the Computer Revolution which has greatly extended, and is continually extending, control over the processes of manufacturing.

Some indication of the tremendous strides that have taken place in the world of computers can be seen from the following:

  One authority has observed that the first computer sold in 1950 filled a room yet it is now possible to buy a micro-computer with a lot more capacity which is 30,000 times smaller. At the same time the cost in 1950 adjusted to present values was £1 million, whilst the equivalent micro-computer today is £200. Because there are no components there is very little to go wrong and thus the possibilities of failures and breakdown which cannot quickly be restored by replacing the relatively cheap silicon chip, are very small. The micro-computer is therefore very reliable. (Management Accounting May 1981)

The job of a computer is to interpret a series of binary digits, which in the earliest computers were represented by cumbersome and inefficient thermionic valves, which were superseded by the electronic switch or transistor. Yet a further development was the integrated circuit, which enables all the electronic components of a whole computer to be contained on a tiny wafer of silicon: the silicon chip or micro-processor. But already a successor is within sight: the Josephson circuit, which will make computing thousands of times faster by utilising the super-conductivity of certain materials at temperatures near absolute zero.

In the fifties the first “automation scare” was at its height. Subsequently, concern subsided only to reappear recently. Last year, for example, a confidential Labour Party policy document captured the headlines by arguing for a “socialist planned economy” (more state managed capitalism) to meet the challenge of the micro-chip.

But what justification is there for this renewed concern? Marsh offers four reasons why “1980-style automation will have a bigger effect on work than the first computer controlled factories”. The first is the growth in the number of computers. In 1960 there were a mere 10,000. Today there are around 700,000 “big” computers costing more than £10,000 as well as tens of millions of smaller ones based on micro-processors.

Secondly, the technical advances in micro-electronics have enormously increased the versatility and scope of computer technology. The significance of the chip is in its size: it is not only cheap and fast (messages don’t have far to travel) but can be used in places where a computer could not be accommodated before. As a result the purpose of computers is no longer simply to store, manipulate, retrieve and organise masses of statistical information. Increasingly they are used in control, monitoring, regulation and automation.

The emergence of the industrial robot neatly illustrates the point. A robot is defined as a mechanical arm which is controlled by a computer. Of the 8000 world robot population in 1980 almost all were “first generation”—big, clumsy, expensive devices used mainly for such work as welding and paint spraying. The newer lightweight “second generation” robots now being produced can manipulate items with amazing precision. An example is the PUMA made by an American company, Unimation, which costs £20.000 and can position things to within 0.1mm.

One reason robots have hitherto made little headway in Britain is the low wages here compared to mainland Europe. As an article in the Guardian explained: “We are in a curiously inverted trap; low labour costs depress the necessity to use advanced technology which in turn decreases the rate of development of technology itself.” (22/11/79). And as the same article pointed out, Britain does not have an overfull order book nor an insufficient workforce to cope with it. Not that the absence of an “overfull order book” signifies that people's needs have been satisfied; true to its absurd and contradictory nature, capitalism is in a slump, intensifying the poverty it compels workers to endure in an age of potential plenty.

But the growing dexterity of robots together with their declining costs in relation to labour costs, make them an increasingly attractive proposition. Indeed, according to statistics from the US car industry, the total cost of the human worker is now $19 an hour while that of a robot is $5.40 an hour (Guardian 19/5/ 81.) And there is another factor to take into account. While automated equipment may work all the anti-social hours an employer may want and not go on strike, it does have a drawback; it cannot be sacked should the state of the market render its loyal services redundant.

To some extent this disadvantage has receded with the advent of the robot. Hitherto, it was mainly the biggest firms with funds to match, involved in the mass production of thousands of items in a single production run, who could afford to automate. The equipment is normally unprogrammable, designed to carry out a fixed routine.

Not so with the robot, which can be programmed to make a variety of products. This means that automation need no longer be restricted to mass production by the economies of scale. Using the same robot—which is within the price range of of even small firms—the automation of small-scale “batch production" has become a distinct possibility, allowing the manufacturer to rapidly switch to a different product should the need arise.

The implications of such flexibility are significant. According to the Economist (24/3/78), “ . . . over four fifths of the world's manufactured products are still made in batch lots of between 10 and 50 units usually at a cost of at least five times that of a mass produced item". The automation of batch manufacture will also be made easier by the chip based computerised scheduling of changes in the raw materials needed.

Moreover the potential scope of robot applications is growing. Recently, the Japanese Industrial Robot Association reported that there are more than 100 products. among them freezers, calculators and TV sets, now assembled by human factory workers that could be made by existing robots which can “see". (New Scientist 12/6/80). As for the future, Joe Engelberger, president of Unimation, has estimated that “. . . the next generation of robots-mobile, multi-armed, understanding speech, talking and with rudimentary vision and a finer sense of touch — would be around for practical use in about three years time.” (Guardian 19/5/ 81)

The impression that may have been gained, that the drive to automate will spell a massive loss in jobs, needs to be put in perspective. Though the increased use of computer technology will entail a certain amount of job loss, its application will open up new areas of work—for example in software (programming). This is shown by the increasing ratio of software to hardware costs which were 2:1 in 1973 and are forecast to be 10:1 by 1985. Mass unemployment is not the result of technological progress but is a feature of capitalism in a depression. The bleak outlook for the world's economy is another reason for Marsh's concern over automation. The growth rate in the seventies was roughly half that of the sixties. This, argues Marsh, “decreased the opportunities for governments to create and pay out of taxation, the jobs in services that did much to absorb the people squeezed out of manufacturing over the past 20 years.”

The suggestion here that automation will serve to exacerbate the problem needs to be examined in the light of the circumstances that persuade or dissuade an employer to automate. Under capitalism the direction, pace and extent of technological innovation is dictated not by human need or conscious will, but by the prospect of profit. Without profit production grinds to a halt, irrespective of the needs of people or the degree of technical sophistication employed.

Another characteristic of capitalism is its basic instability. Unplanned, it moves in continuous cycles of market expansion and contraction. Though a boom cannot be sustained indefinitely neither is there such a thing as a permanent slump. Marsh observes that the effects of new technology on the labour force will be far less marked when it is introduced at times of increasing demand for goods than when demand is falling. But in a recession when much existing technology is often underused, the incentive to introduce new technology will be considerably less anyway than at a time of rising demand, when firms strive to increase output and offset the increased cost of labour power.

If the decline in manufacturing industry is irreversible, runs the argument, future employment would have to be found increasingly in the service industries if mass unemployment is to be averted. Ironically, the pessimistic view of the effects of new technology on work is centred more on service than manufacturing:

  The information occupations (secretaries, typists, clerks managers) are thought to amount to 65 per cent of the working population so that even moderate improvements in productivity could bring about unemployment levels in the region of 10 to 20 per cent unless offset by compensatory increases in demand in these and other activities. By comparison, the impact of new technology on employment in manufacturing is expected to be relatively small. The overall consequences are said to be comparable with the industrial revolution. (New Scientist 26/4/79)

An example of “improved productivity" would be the truly paperless office, a vital requirement of which would be a practical electronic method of filing documents quickly and cheaply. Philips have already developed a prototype system called Megador which achieves the necessary speed—it can read and store an A4 page in one second and retrieve any document in 5 seconds—but is regarded as too expensive for all but the largest firms (New Scientist 23/4/81) Twenty years ago in America Congressman Adam Powell launched an official investigation into “the impact of automation on employment” with a warning that “the Frankenstein of automation has come closer to us”. If ever there was a case of mistaken identity it is this. That the elimination of drudgery and the increases in productivity that automation makes should come to be seen as a threat is an indictment of the society we live in.

Machines are neither masterful nor submissive. It is men and women, having or lacking possession of the means of wealth production, who relate to each other in this way. Capitalism is the monster we ought to fear and hate. Who can regard a system that perverts technology into a means of inflicting global destruction with anything but apprehension? Who can say that such a system is efficient, when so much of its productive effort does not satisfy the needs of people?

Thanks to the computer revolution we are potentially better able now than ever before to organise effectively the socialist production and distribution of wealth. All that stands in the way is an increasingly outdated social system.

Robin Cox