Face recognition: a less-bad option?

Here is a very nicely-constructed essay by Jane Bambauer, a Professor of Law at the University of Arizona.

“This essay”, she says, “does the unthinkable — it defends the police use of facial recognition technology to identify suspects in crime footage or to locate individuals with outstanding warrants.”

It’s a well-thought-out and very readable piece, and some of her key arguments are along the following lines:

  • We currently have very harsh punishments for relatively minor crimes (especially in the USA). This high level of incarceration is not the best way to deal with the problems, especially since the success rates for rehabilitation are so low.

  • We do this at present, though, because the crime detection rates are so low that it’s important that the penalties are very high if they’re to act as a disincentive.

  • A much better and more progressive route is to detect much more crime and punish it less severely. This has been shown to be a much better disincentive, too. But technology is key to achieving any significant improvement in detection rates.

  • Facial recognition technology is an important tool here and, though it has been shown to have problems with bias etc, it may actually be less biased than other forms of surveillance.

You may or may not agree with the above, but if you’re interested, it’s well worth reading the 9-page article before jumping to conclusions. (Many thanks to Charles Arthur for the link.)

So here’s a related question: would our roads be safer if everyone was fined a small amount every time they broke the speed limit, rather than receiving a severe penalty on the rare occasions when they were caught? I think the answer is clearly ‘yes’. Would you be willing to put up with having your speed monitored all the time in order for this to happen — as long as everybody else had to do the same, of course?

Enjoyed this post? Why not sign up to receive Status-Q in your inbox?

1 Comment

Thanks Quentin!

A friend of mine was recently telling of a pay-as-you-go car insurance that he has bought.
He does very low mileage and just got a car because now he lives in the suburbs he needs to get to the shops occasionally.

The one he bought takes the form of a dongle that attaches to the car’s diagnostic port and a phone app that tells him what he’s paying.
According to the Ts&Cs all the dongle does is check the mileage and there was quite a lot of language about how they definitely don’t use it for anything else. For example, seeing how you are driving or as evidence of fault in an accident.

…but, as with other car automation things such as automatic breaking that can’t be disabled and therefore sometimes gives you advantageous car insurance premiums, there seems to be a lot of scope for innovation and new uses here.

How would you envisage applying you idea? As a mandated thing or as a way to get better insurance premiums? The latter seems more acceptable to the market but might be difficult to implement commercially because they’ll end up with lots of evidence of law-breaking that they couldn’t afford to disclose.

Got Something To Say:

Your email address will not be published.

To create code blocks or other preformatted text, indent by four spaces:

    This will be displayed in a monospaced font. The first four 
    spaces will be stripped off, but all other whitespace
    will be preserved.
    
    Markdown is turned off in code blocks:
     [This is not a link](http://example.com)

To create not a block, but an inline code span, use backticks:

Here is some inline `code`.

For more help see http://daringfireball.net/projects/markdown/syntax

*

© Copyright Quentin Stafford-Fraser