Prediction, judgment and digital technologies

He had been told that when looking for a good oracle it was best to find the oracle
that other oracles went to, but he was shut. There was a sign by the entrance saying,
‘I just don’t know any more. Try next door, but that’s just a suggestion, not formal
oracular advice’.^[1]

The Web Summit has just finished, and as expected there’s been no shortage of future-gazing,  prediction and reinvention.  It was held in Lisbon, a city that, as the Guardian put it, is currently ‘reinventing itself as Europe’s leading tech hub’:

Even if the comparisons with San Francisco are fanciful – and, frankly, extend little beyond the fact that each has a red bridge, hills, trams and good surfing – Lisbon has moved significantly closer to fulfilling its ambitions by winning the race to host this year’s Web Summit.

Back in 2007, in my book Transforming Legal Education, the Afterword winds forward to 2047, and the heroine, Anna, is heading off from her law programme in Glasgow and the virtual town of Ardcalloch at the University of Scotland for a six-month placement in Lisbon, attracted by the improved weather and maybe also the law/tech opportunities.  So my prediction of Lisbon as a tech centre was right, but the timing was well out.  Of course no one predicted Brexit either, rather a large spanner in the works, for Anna in 2047 and for us now in 2016.  Though I remember wondering even when I was writing the Afterword if an independent Scotland might not have better HE possibilities in Europe than hitched to a disUK that wasn’t sure what it wanted in or from Europe.  Retrospectively, what a wasted opportunity Indyref1 was, back in 2014, in spite of the economic difficulties it would have caused.  They look short-term now, compared to the long-term decline that’s predicted for Brexit Scotland (and that includes the substantial impact on Scots HE).

After Lisbon I predict (‘formal oracular advice’…) Edinburgh will be the next tech centre — the Scottish government has to do something to attract new talent into the Scotland after Brexit.  Like San Francisco it’s got trams, hills, bridges (George IV Bridge etc); and the culture of Scottish medieval & Enlightenment architecture must surely influence the code.  Surf – yes, well the creatives would have to load up the VW van and head west to the Outer Hebrides, or north, out beyond Orkney, far beyond Shetland, to the truly epic winter Atlantic off the Faroes.  The lure of the pre-modern, the fabulous north may grip them as it has me, sparse as code and saga.  They may not return.

Prediction is on the rise as the digital saturates our society.  The past is as difficult to predict as the future, it would appear, but neither is entirely impossible given the proper metadata and the proper approach to its collection and interpretation. We see that in the rise of machine learning, which has now largely overtaken classical AI approaches to metadata (I’m in the scruffie camp, tagging along behind Minsky, Schank & Winograd), whether it’s the prediction of judicial decisions, or the shaping and predicting of a US presidential election outcome, as this article from Wired points out.  In the Editorial [paywall] to the special issue of The Law Teacher that I edited earlier this year one of the predictions I made was that augmented intelligence would be one of the key developments in the next decade — and as this article from Inside Higher Ed points out, IBM is making practical steps in that direction, picking Blackboard and Pearson to platform Watson in the adaptive and personalized learning markets.

But essential to prediction, and to the construction and interpretation of predictive data in any field, are acts of judgment.  When Flight 1549 ditched in the Hudson River after bird strikes knocked out both engines while taking off from La Guardia, counterfactual questions were asked in the aftermath.  Was the ditching necessary?  Couldn’t Sullenberger and his flight deck crew have made it back to La Guardia, or to Teterboro airfield in New Jersey?  The data was plugged into Airbus flight simulators.  What would it predict about the past?  Here’s Wikipedia’s account of the result:

The test pilots were fully briefed on the series of events and maneuvers. The test pilots were able to return successfully to either airport in only eight of 15 attempts, although four out of four attempts to reach the closest LaGuardia runway were successful. The NTSB report noted that these test conditions were unrealistic: “The immediate turn made by the pilots during the simulations did not reflect or account for real-world considerations…” A single follow-up simulation was conducted where the pilot was delayed by 35 seconds: he crashed trying to return to LGA runway 22.

Decision-making in the moment, in the critical moment, is the heart of the matter; and no matter how deeply technical, it is also profoundly human.  I remember reading about the First Officer racing through pages of protocols designed to be checked at 35,000 feet, not 3,000 feet, and trying to imagine how he managed to do that.  Training, education, habit, coolness of mind, courage, attentional focus, shifting expertly from past to future in the present moment, selfless thinking, detailed thinking in the context of utterly new conditions (eg the pitch of the aircraft when striking water) – even as we count up the qualities that were required by the flight crew we’re out of the moment.  The test pilots conducting the sim were also out of the moment – ‘the test pilots were fully briefed’ – and of course they’d read the newspapers so they all knew the outcome.  As a result, and as the National Transportation Safety Board observed, they could never be in the position of the original crew, who had no foreknowledge of the event they were in, nor retrospective knowledge, who knew desperately that seconds counted as the aircraft fell towards New York, but were uncertain how to make them count for the best and, facing death, just didn’t know how this was all going to turn out for them and their crew members and 155 passengers and possibly many others — which in the cool language of the NTSB is rather a crucial ‘real-world consideration’.

What those qualities and no doubt others supported was not a calculation – data input, data output – but a process of experiential, experimental discrimination, a form of judgment.  Apart from the act of love, the act of judgment is one of our finest qualities, which is why love and judgement are so intimately linked across the Arts, in our histories, drama, films, literary narratives, visual arts, music. Their interaction is a central theme of fiction, films and TV dramas on the rise of robots and AI.  And they are foundational, I would argue, in how we understand and shape digital technologies for education.

They are also foundational for our students, now; but we do so little in our law programmes to enable them to understand and shape digital technologies, for their own education and that of others.  Nor do we help them understand the profound effects that digital is having on society in general, and not just on the economy but on how it is redefining qualities such as professionalism, and the effects it is having on our personal and emotional lives.  Back in 2007/2047 I predicted that some students would form emotional attachments with the personal digital assistants that assisted their learning, some falling in love, others disliking them.  Given the work of Turkle and others, which has since taken a dystopian turn, that was an easy prediction.  Anna, my heroine of the Afterword, is too self-aware to fall for that.  She’s learned from the Foundation DL (Digital Learning) courses at school and the USG (University of Scotland Glasgow) all about the effects of digital on human emotion, on scholarly activity and on professionalism. It’s just part of her life, in the same way that she swiftly, supported by her exocortices, learned Portuguese in preparation for her time in Lisbon (Brazil beckons, next).  And learned to surf.

1. ^[1]Adams, D. (1993) Mostly Harmless.  London, Pan Macmillan, 73.↩