Planet a.f.p.

One could suggest that my attitude to big – or rather, huge - databases of personal information is a bit schizophrenic. In fact, I suggested it to myself, and came up with a justification which satisfies me at the moment. Others may, of course, differ, and I would be glad to hear other opinions.

On the one hand, we know that Google, Amazon, Facebook and their peers are gathering as much information as they can about us, with the aim of cross correlating it all and passing on knowledge about our preferences to those enemies of mankind, advertisers. And I have been quite relaxed about that. Firstly, I see it as inevitable, and therefore there is little point in wasting effort in a struggle which we cannot win. There simply is not enough popular will to fight the commercial pressures of these advertisers. But also, there is some truth in the claims by these companies that their targeted advertising is beneficial. If well targeted, and the ads are, as on Google, reasonably unobtrusive, there is actually a win-win situation here. The advertisers pay for a service which I get to use for free. This has been true for commercial television (and radio before it) for decades. But the computer-based targeting of the ads means they are much more appropriate to me, unlike ads on TV which are generally very crudely targeted. Which is better both for me and for the advertisers – increasing their willingness to pay for the service I use. I am not totally relaxed about it – there are dangers which I will come back to – but on balance I find the benefits outweigh the dangers, especially if we note and take precautions against the dangers.

On the other hand, I have been vociferous in my opposition to the last government's (thankfully deceased) plans for a National ID Card and the associated intention to provide a unique index to hundreds of government, and other databases – with the possibility, one day, of merging them. Some of my objections have been at the nerdy level – the system would never have worked as designed, and it was a blank cheque for the taxpayer to be ripped off by IT snake oil salesmen. And some of it was about the number of people who, precisely because of the huge variety of data held on the system, would have access to and would be able to leak it. And also the threats to civil liberties represented by the system in the hands of potentially repressive future government. But I also have a much more visceral, emotional dislike of the database. This emotional dislike prompted me to look closely at the whole scheme, and drove me to discover significant numbers of very real problems with the system. The problems were always there and always real and huge, but I would not have been prompted to look at them without my emotional distaste for the system.

So why the difference in my responses to the two systems? They are both accumulations of huge amounts of data about me. Some of that I gave intentionally, some unintentionally. One might expect that I would be happier with the government database, over which I might theoretically have some control through my elected representatives (though theoretically pigs might fly with suitable rocket assistance). Much better, surely, than the Big Corporations, based in foreign countries, over which I have effectively no control?

Part of it is the content. Google and Co. are collecting data which is, item by item, probably innocuous. Even if they collect more than they are now – perhaps getting positional information from my mobile phone and purchase records from my credit card – no single piece of information is that worrying. The phone company needs to know where I am, the bank needs to pay my charges: this is information I have given voluntarily (even if, in the case of some people, it is unwittingly). It is in the correlation that danger arises. Tracking my position maps my movements, my pattern of purchases may tell tales about me. And correlating those against others may tell about who I meet ans so on. Whereas the government database may have single facts I don't want to get out – my medical and police records, for example. But, important though this is, it is the source of a logical fear rather than my emotional dislike.

The solution I have come up with is that what matters is the reader of the database. The “reader” of Google's database is (or should be) only Google's computers, trying to correlate all the data it has about me with that which it has about other people. The simplest example is Amazon's “people who bought this also bought...” - but the technique can be taken far, far further. And all the big Net companies are labouring furiously to do so, which is the cause of much fear in some people. But not in me, because I trust computers. Not to be fair, honest, or truthful. But I trust them not to think, and not to make judgements.

By contrast, the ID Card complex of databases is designed to be read by humans. There may be sundry automated searching and correlating mechanisms, but they are designed at the end of the day to spit out information for humans to read. And humans think – and humans judge. And humans are prone to all sorts of misjudgements. They can be hasty, they can be over-influenced by single items, they can have value systems totally different from mine.

To take an example, supposing I accidentally stray onto a truly loathsome website, the fact that I have done so will, in all probability, stick in the mind of a human reading my web log. To those searching professionally for the purveyors of nastiness, this will lodge in minds trained to think the worst. But for a computer correlating my tastes against those of others, it will be the tiniest datum. It will be correlated with the hundreds of other sites I see, and compared with the patterns of other users, and no match will be made if I only visit the once.

And looking at the other side – supposing I do have unconventional tastes. The software will blindly match my pattern of web access (and, in the larger databases, purchases and visits to geographic locations) to those of other people with similar tastes – and offer me things they like. It won't judge me. It won't say, or even think, that I am a loathsome person for having such tastes – because, as a computer, it cannot think. If, given a large enough database, it discovers I have certain tastes, it will blindly serve those tastes, to my benefit. And the larger the database it has, the more likely its blind pattern matching is to find what I really like, even if I don't realise it myself, and deliver adverts for or pointers to that thing.

Now let us turn to the other database. It delivers from its digital guts facts about me – to people. Let us forget about the possibility of faulty data (huge though that possibility is). Somebody knows something about me. A relationship has been established between us. But it is without my will, without my knowledge, and it is a strictly one-way relationship – they (I even have to use the false plural, because I don't even know his/her gender) know about me, but I don't even know they exist.

And, as I have mentioned in many other contexts, humans are above all a social animal – Homo Socialis. We define ourselves to a large extent by our relationships. And we like to control those relationships. When we meet someone, we want the opportunity to show our best side. And to respond to their reaction to us – to tune our behaviour to make us look as good as possible in their eyes. We all want to be liked. If we can be looked up to and admired, so much the better. It is a game we play, whether consciously or not, every time we meet someone. We tweak and tune our behaviour so they get the best view of us. Of course, most (but not all) of us are still fundamentally honest – we present a truthful view of ourselves, but one tuned to put us in the best possible social light. After all, you can always drop your standards later if you have over-stated your case, but it is much harder to pull up when someone has seen an unflattering side of you.

And yet this person you don't know – or rather, these people, of names, ages, sexes, and attitudes unknown to you, have got the drop on you. They know things about you, and you know nothing about them. It doesn't have to be important things they know, though it is worse if they are. But they have robbed you of your right – and I thing we do emotionally see it as a right – to present your best side, to respond to their implicit (or explicit) criticisms, and to make the same judgements of them as they are of you. You have been sent into the jousting field of social relationships naked, shackled and blind. And you (or at least I) really don't like it.

This rather clearly informs where the line should be drawn by the big data gatherers. Where people get their hands on identifiable data. As long as it remains in the purview of the computer alone, it doesn't matter. If its correlations show that I like, for example, something called “blue cheese”, that doesn't matter. The computer doesn't know what blue cheese is – even though it could probably list hundreds of different brands of it. It can offer me blue cheese related websites, send me blue cheese newsletters – I will probably like them. It also do it for other people, as long as identity is kept out: if it tells some medical researcher that a taste for blue cheese correlates with early deafness, it will probably be to my benefit. But let it come before the eyes of human who can correlate my tastes with me – even me as a de-identified database id – and a line has been stepped over. A blue cheese hater might think the worse of this unknown person who can consume, even search out, the devil's food. And, even if they never find out who I am, by having an image in their mind of me they had put the first toe on the slippery slope. Which does not happen with the brainless computer.

Which leads to another thought on an apparently totally unrelated subject, the burqa (and the niqab), the face covering robes worn by a small minority of Muslim women. The response to these robes seems to partake of the same visceral loathing as my response to the government databases. I think this is for the same reason.

I should make one point clear. These robes are often associated with a cruel and controlling dominance of women by the men of their community. Such control is unacceptable to our society, and the women should be supported in every way possible to break that control and live their own lives. But the burqa is a symptom of such control, not a cause of it. It is the control which is to be fought. If the women were free to make their own choices, they could remove the offending garments – assuming they wanted to. But it is clear that there are at least a few women for whom the burqa is their own choice. These women leave many of the opponents of the garb squirming uncomfortably. On the one hand, if they are wearing them of their own choice, our very desire that they should make their own choices says that they should be as free to wear the burqa as they are not to. On the other hand, the visceral dislike is still there. Opponents are left mumbling that it is still symbolic of that oppression even if the oppression is still there.

I suggest that it is much simpler: the face-covering, emotion-hiding burqa gives the hooded woman something of the same “unfair” advantage held by the hidden reader of our digital records. They can see our response to them, they can “read” our faces, but we cannot do the same to them. And that subliminally violates our feelings of fairness. And precisely because it is subliminal, unanalysed, it is difficult to put into a logical context.

Make no mistake, I abhor the oppression of women in certain Muslim cultures, and agree that it should be ended, But fighting the burqa is fighting the symptom not the disease, and is a distraction from the real fight.