Nadine Dorries’ proposed amendment to abortion legislation, stripping termination clinics of their counselling role, and bringing in third party counsellors revealed as American rightwing ‘Christian’ beard boys and hockey mums, has been heavily defeated in parliament. A sigh of relief perhaps, but not for long. Look at the undercurrents of this debate. One can hardly be surprised that the three ministers whose support she managed to gain were penis-headed Iain Duncan Smith, work and pensions secretary, tricksy Liam Fox, missile (penis) secretary, and Owen Paterson, Northern Ireland secretary, whom I admit I have never heard of. Well and good that their secretions have been flicked into the dustbin.
Earlier this week I attended a meeting about a certain proposed public health information project which seeks to shoot and digitally host a number of video clips explaining various aspects of pregnancy, labour and early care. Fair enough, we are an increasingly atomised and isolated species, there just ain’t the friendly mums and aunties and sisters there used to be…and they might not give the right advice anyway, so better look up stuff about babies on a computer screen rather than ask someone.
But then there’s the question of…well, what are the questions? We know that women ask questions about pregnancy and labour and babies…but what are they asking? Those who know how Government works will know where this goes…
Enter a penis-headed man in dark suit armed with a laptop containing the output of survey data fed into a NLP engine – no, not Neuro Linguistic Programming, but that old Artificial Intelligence discipline of Natural Language Processing. Now it is possible to process survey responses in lightning time, with NLP syntactical/ statistical analysis, and, if some virtual black box referred to in this man’s presentation as ‘machine learning’ (clipart icon of a calculator was used) is to be believed, an analysis of sentiment is also performed.
Sentiment? Do they mean emotion?
We could spend pages and pages here examining the veracity of machine learning algorithms, ultimately no doubt having to delve into Noam Chomsky’s Universal Grammar theory (and refutal), the Strong Symbolic AI of Daniel Dennett and others, modern neural net theory and all sorts of difficult mathematics. We could debate whether symbolic analysis – for that, surely, is all an algorithm, no matter how sophisticated, running on Windows 7 on an Intel chipset on a standard laptop architecture inside a plastic case could ever do – can encompass the watery realms of emotion. (In my experience water is highly detrimental to electronic devices.)
Or we could simply do what the Government is evidently unable or unwilling to do – appeal to common sense. To wit – what has happened to us that the question of what questions women ask about pregnancy, labour and babies, is to be answered by a man with a laptop? Or if it is that the friendly mums and aunties and sisters and midwives and GPs and health visitors are no longer there, then is that not what we should be looking at?
Back to the machine learning for a moment. What did it learn? That the top question in women’s minds is breast feeding. Now, even I could have told you that.
And it’s tempting to leave the argument there, but we must reconsider. What did the machine tell us? Breast feeding. But what did the machine learn? And here we visit the murkier side of all this, the side that French philosophers Paul Virilio and Jean Baudrillard flagged up years ago before even Matrix came out. This is the flip of the computational metaphor (where, as means of structuring enquiry into Mind, it was considered as as computer, as a machine). That, the more we turn to machines ofr answers, the more machine-like we become.
And they are already here, those metal dragons. In sex, in breastfeeding, in labour, in feeding, in babies. As Virilio puts it, an ‘unspeakable contamination’.