Black boxes

Interesting snippet about the Face ID system in Apple’s iPhone X.

Even Apple will not be able to explain how its forthcoming iPhone X can spot some efforts to fool its facial recognition system.

The firm has released a guide to the Face ID system, which explains that it relies on two types of neural networks - one of which has been specifically trained to resist spoofing attempts.

But a consequence of the design is that it behaves like a "black box".

Its behaviour can be observed but the underlying processes remain opaque.

Really this isn’t so different from the OCR systems I was involved with over a decade ago. If the OCR system read flow as flaw, nobody would ever know for certain why. We’d pick it up with error correction systems and move on. I doubt if those who wrote the underlying recognition software could tell us in any but general terms.

These machines “know” what they are doing but we don’t know in detail and they can’t pass on that detail because they would hit us with impossible amounts of information. So as far as possible we let them handle it and only get involved when things go wrong, which probably isn't as often as we'd like to think.

It isn't easy to dwell on modern technology without feeling somewhat superfluous. Are we here to give the machines something to do or is it the other way round?