Tuesday, 14 March 2017

Automation, Work and Education


Our world is becoming increasingly automated and this increase appears to be having an effect on the number of jobs available. It is possible that in the future automation might not only lead to a decrease in the number of existing jobs but also create an increasing number of different jobs. A second possibility is that automation will mostly lead to a decrease in the number of jobs. In this posting I want to examine some of the consequences this second possibility has for work and education.

Pessimists might argue that a widespread loss of jobs will lead to widespread hardship and poverty. I believe such a pessimistic outcome is unlikely because such an outcome would threaten the survival of both the state and the market economy. In this situation both the state and the markets would have reasons to introduce some form of universal basic income, UBI. According to Tim Dunlop UBI means,

“A basic income, on the other hand, is the idea that everyone should be paid a minimum monthly income that allows them to meet their basic economic needs.” (1)

It is important to note that UBI in response to increasing unemployment caused by automation is not some attempt to reform the benefits system but rather an attempt to counter an existential threat which might be posed to the state due to this unemployment. It might be speculated that UBI might not just be useful in combating the effects of unemployment but might also be necessary for the continuation of capitalism. In an age of large scale automation, capitalism might survive without workers but it seems doubtful if it could survive without consumers In the rest of this posting I am going to assume that if automation causes widespread job losses in any state that that state will introduce some form of UBI in order to counter this existential threat. I will further assume that UBI will be large enough to permit people to live in moderate comfort.

Some might think that automation and UBI will lead to some golden age. In the ancient world the upper classes in Greek and Roman society led a life of leisure in which most of the work was done by slaves. It might be argued by analogy that automation might introduce a golden age in which we live a life of leisure with most work either becoming automated or done by robots. I believe such a golden age is an illusion for two reasons. First, upper class Greeks and Romans may have lead happier lives than their slaves but there is no evidence that they lead happier lives than people living now. The ancient golden age at least for some appears to be an illusion and so any argument by analogy fails. Secondly if we live in a world in which all the work is automated or done by robots we might suffer from the unbearable lightness of simply being. We might feel our world has lost all purpose and that we simply exist. We might become bored. Limited boredom might encourage us to take steps to alleviate our boredom but prolonged boredom is harmful. According to Harry Frankfurt boredom is not some innocuous discomfort but something that threatens our psychic survival. (2) I have previously argued that a world whose inhabitants are bored and feel they are simply existing is a dangerous world, see riots and the unbearable lightness of simply being . It is possible that even if automation frees people from work and that the resultant widespread loss of jobs does not lead to widespread hardship and poverty that it might also lead to people’s lives being degraded rather than some golden age.

The above pessimistic scenario seems to be a realistic possibility and I now want to examine what might be done to counter the negative effects of such a possibility. Prior to my examination I want consider what we mean by work. Work might be roughly defined as making an effort for some economic reward or hope of such a reward. However, such a definition is at best an incomplete one. I have suggested previously that someone might work in her garden purely for the pleasure it brings her without any thought of economic reward. Hannah Arendt suggested there is a difference between work and labour. According to Arendt labour is what we do in the normal process of living in order to survive. For Arendt work might be simply defined as any human activity which is not driven by our need to survive. Arendt’s definitions are interesting but also seem to be incomplete ones to me, dancing is not working. Intuitively work requires some effort. Work might be now defined as any human activity requiring effort which is not driven by our need to survive. Such a refined definition also seems an incomplete one. If I am running away from a bull I might make a great effort but I’m not working. Work might be now defined as any human activity which matters to us requiring effort which is not driven by our need to survive. I believe Arendt’s insight is important and I will use it to define two different ways of working. I believe it might be better to label labouring as ‘working for’ something we need to survive. ‘Working for’ something has mostly instrumental value. Work defined as a human activity which matters to us requiring effort which is not driven by our need to survive might be labelled as ‘working at’. ‘Working at’ has mostly intrinsic value.

Let us now examine the possible effects of increasing automation bearing in mind these two definitions of work. Let us assume that automation might decrease or even eliminate our need to ‘work for’ things, to work instrumentally. Does this decrease matter? I would suggest it does matter to someone if she doesn’t ‘work at’ something. In such a situation it seems highly probable that such a person might suffer from the unbearable sense of simply being. She might feel her world has lost all purpose and that she’s simply existing. It follows we have some reason to fear the effects of increasing automation.

Assuming we aren’t Luddites and don’t want to or can’t stop the progress of automation what steps should we take to mitigate some of the worst effects of not ‘working for’ anything? First, if automation greatly decreases our need to ‘work for’ we would need to refocus our education system. At the present time at lot of education focusses on equipping people for jobs, to ‘work for’. Let us assume people no longer need to ‘work for’ and that a purely hedonistic lifestyle also leads to a lightness of simply being. In such a situation ‘working at’ something might help counter someone’s sense of simply existing due to her ceasing to ‘work for’ something. In this situation education should focus on enabling people to ‘work at’. In order to do so science education remains important because we need to understand how the world we live in works. But we also need to simply understand how to live in such a world and to enable us to do so education should place greater emphasis on the humanities.

I have argued in a highly automated age people need to become better at ‘working at’ something. All work can be good or bad and this includes ‘working at’. Someone might ‘work at’ doing crosswords all day. I would suggest this is not good work. If ‘Working at’ is to replace working for it must be good work. Samuel Clark defines one element of good work is that it requires some skill. According to Clark,

“the development of a skill requires: (1) a complex object and (2) a self-directed and sometimes self-conscious relation to that object.” (3)

I now want to consider each of these requirements. According to Clark good work involves working at something which must have some complexity. According to Clark the something we work at must have a complex internal landscape of depth and obstacles (4). He gives as examples of a skilled activity, music, mathematics, carpentry, philosophy and medicine. Doing crosswords might be a difficult task but it lacks complexity. Clark argues good work must be self-directed. Let us assume someone is self-directed to work at some complex task purely to mitigate her sense of simply being. I would suggest that such self-direction fails. Why does it fail? It fails because in order to prevent this sense of simply being someone must work at something that satisfies her. For an activity to satisfy someone she must care about that activity. Let us accept that Frankfurt is correct when he argues ‘caring about’ is a kind of love because the carer must identify with what she cares about. It might be concluded that good work is doing something complex which the doer ‘cares about’ or loves. It might then be suggested that provided people can ‘work at’ something and that this is good work and that this ‘working at’ might mitigate the some of the effects of job losses due to automation.

 However even if we accept the above difficulties remain. Let us assume any good work either ‘working for’ or ‘working at’ requires some skilfull action. Let us further assume a skilful action requires that the doer must identify with her actions by ‘caring about’ or loving them. Unfortunately, ‘caring about’ or loving is not a matter of choice.

“In this respect, he is not free. On the contrary, he is in the very nature of the case captivated by his beloved and his love. The will of the lover is rigorously constrained. Love is not a matter of choice.” (5)

It further if someone simply chooses to ‘work at’ something in order to compensate for her loss of ‘working for’ that this ‘working at’ need not be good work and as a result won’t mitigate her sense of boredom. Someone cannot simply choose to do anything to alleviate her boredom. If she simply chooses it seems probable her choice will bore her. She must ‘care about’ what she chooses. If society is help mitigate the effects of job losses, due to automation, then it must create the conditions in which people can come to care about doing complex things. I have suggested above that education might help in this task. W B Yeats said ‘education is not the filling of a pail, but rather the lighting of a fire’ perhaps education must fire peoples’ enthusiasms every bit as much as enabling their abilities. Perhaps also we should see learning as a lifelong process. Lifelong education broadly based which fires peoples’ enthusiasms might help create the conditions in which people can ‘work at’ things hence mitigating some of the harmful effects of job loss due to automation.


Lastly there are activities which might mitigate some of the harmful loss of jobs which have little to do with work. Music and Sport would be examples of such things. Of course it is possible to ‘work at’ music and sport, we have professional sportspersons and musicians, but most people just play at such activities. Play is a light hearted pleasant activity done for its own sake. Play is important; especially for children. It might be suggested that some forms of play are a form of good ‘working at’. All work is goal directed and so is some play. Perhaps there is a continuum between work and play with the importance of the goal varying. Perhaps in an automated age play should become more important to older people also. Activities playing sport or music require some infrastructure and perhaps in an automated age it is even more important that society helps build this infrastructure. At the present time governments foster elite sport. Perhaps this fostering should change direction to fostering participation rather than funding elite athletes.

  1. Tim Dunlop, Why the future is workless, (Kindle Locations 1748-1749). New South. Kindle Edition.
  2. Harry Frankfurt, 2006, The Reasons of Love, Princetown, University Press, page 54
  3. Samuel Clark, 2017, Good Work, Journal of Applied Philosophy 34(1), Page 66.
  4. Clark, page 66.
  5. Frankfurt, 1999, Necessity, Volition, and Love. Cambridge University Press. Page 135.

1

Engaging with Robots

  In an interesting paper Sven Nyholm considers some of the implications of controlling robots. I use the idea of control to ask a different...