A is for Asimov; and there are two in Big Data

Asimov’s big idea

With the current interest in AI and where it might be leading us it is no surprise that the name Asimov is frequently referenced in discussions. Isaac Asimov, although a scientist and educator in his own right, is best known as a writer of Science Fiction. In particular, his stories about a future society grappling with how to assimilate robots seem very prescient today. In these stories he introduced his ‘Three Laws of Robotics’ which have become so well-known that people often quote them as if they are actual scientific laws on a par with those of Newton:

1.     A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2.     A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

3.     A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.

 

 These laws are written into a robot’s software to control their interactions with humans. Much of what Asimov wrote as science fiction is now science fact; for instance in the opening chapter of his Foundation series he describes some-one using a pocket calculator and another character reading information from a VDU screen. Unremarkable you might think, until you realise these stories were first published in the 1940’s. The Foundation novels were set in a future 20,000 years after the Robot Novels. At the end of “The Robots of Dawn” the Robots themselves conclude that their existence is damaging humanity by removing self-reliance and initiative and therefore encourage a wave of space exploration that does not use robots. The Foundation stories therefore describe a society that is Robot free.

Asimov’s other big idea

They also describe a future where humanity pervades the galaxy, living on over twenty-five million populated worlds. This epic stage allows Asimov to introduce another big idea: the science of Psycho-history. This is the statistical analysis of human behaviour through the means of mathematical equations. Developed by the scientist Hari Seldon, Psycho-history was a synthesis of information of a mass of variables brought together through mathematical equations. Asimov looked at the Economics and Sociology of his time and imagined a blend of them, underpinned by mathematics, able to map out the future direction of society.

Just as his three laws underpinned the functioning of robots, the scientist in Asimov created assumptions which were necessary for Psycho-history to be valid. The first assumption was that the statistical sample being studied was large enough to include all variables. The second was that the humans being studied were unaware of the fact so that their behaviour was not altered by the fact of that observation. In the later Foundation stories Asimov explores what happens when people move from observing to manipulating, and what happens when people resent that manipulation. The Seldon Plan became a blue-print to predict the future that underpins the entire series (it nearly undermined the whole series by removing the element of suspense; Asimov had to subvert the Seldon Plan in order to re-introduce jeopardy to his characters lives, but that’s another story).

 

And so to Big Data

At this point you will be beginning the make the same connections I did; that this sounds a lot like Big Data. The inter-connected world we have gradually moved into over the last twenty years has created the opportunity to conduct studies with a view to prediction. These can provide answers to questions such as: When will energy demand be at its highest? when will traffic flows present a risk of gridlock? and so on, with a much higher degree of accuracy than previously possible. The biggest data sample of all is of course made up of the users of Facebook, which whilst nowhere as great as the population of Asimov’s Galactic Empire, still tops out at over 1.6 billion humans.

To manage all this data Facebook, along with other digital giants like Google and Amazon, utilises its own equations, the famous (or perhaps infamous) algorithms. Analysing the buying habits, location changes, demographics, likes and preferences of two billion people is going to throw up some valid conclusions for those who take the trouble. How society came to be persuaded to provide the information is a topic for another article, as are the ethical questions raised by such data harvesting. We are though perhaps at the point where people are beginning to be aware of being manipulated and possibly to resent it. We are slowly becoming aware of the extent to which our personal data may have been used to target us with messages with a view to influencing our voting decisions.

Government by Big Data?

Leaving aside such discussion, perhaps for another post, the point to be made here is that with Big Data we are at least in the embryonic stages of creating a society run on Psycho-historic lines, where policy is dictated by algorithm. Any-one who doubts this should consider the UK government briefings on the progress (if that is the right word) of the COVID 19 pandemic. The press conferences largely revolved around a series of graphs with commentary on direction of travel of the lines on the graphs, how actions affect the data and, finally, how the data will influence future policy. This makes sense, reflecting Sherlock Holmes assertion that it is a mistake to theorise without data. However, this way of making policy has a potential flaw at its heart; namely that in the litany of numbers it becomes possible to forget that each number represents a person; in this case a suffering or even dying person. Numbers can also be spun (there are lies, downright lies and statistics). There have already been examples of this. One scientific advisor ruminated at one point that 20,000 UK deaths from COVID 19 would be “a good result” compared with earlier forecasts that were much higher. I cannot escape a nagging suspicion that by setting out the shape of a graph as the goal of policy the Government is essentially saying that a certain number of deaths are acceptable as long as they don’t all come at once.

Resistance to accepting that data rules runs through the early episodes of the Foundation series. At one point, when the Foundation has finally come out on top in a particular struggle, one character reminds another that the “inevitable victory” they speak of has only come about after months of struggle and the expenditure of many lives, suggesting that human agency is still the determining factor.  

Whilst the Government stance of being guided by the science and the data is understandable they need to be aware of the risks of over-reliance on this, in particular, fooling themselves into thinking that being guided by data absolves them of any responsibility. Moreover, Data can be interpreted in different ways and no-one has a monopoly on wisdom. Data is merely information. It is only useful when human intelligence is applied to it to draw conclusions. As the poet T.S. Eliot puts it, in an uncannily prescient couplet:

“Where is the wisdom we have lost in Knowledge?

Where is the knowledge we have lost in information?”

So, if you have never dipped into the writings of Asimov, especially the Foundation series- both the earlier trilogy and the later sequels, Foundation’s Edge and Foundation and Earth, perhaps now is a good time to do so.  They will provide food for thought on Big Data and determinism as well as being great stories in their own right, whilst if the society he describes seems an awful lot like our own you can reflect on how prophetic he was (to any-one who doubts the connection between science fiction and scientific advance I would say, just take a good look at your mobile phone, then watch an episode of the first Star Trek series. Star trek communicators to mobiles, I think we can all trace a direct line there). For me the appeal of Asimov’s writings is that he makes you think, not least about the impact of technology on society.