When the protests that followed Iran's presidential election in 2009 began to fizzle out, the state-controlled media put photos of the demonstrations online, with the faces of unidentified troublemakers highlighted. Viewers were asked to help identify them and even track them down. If the Iranian authorities are to be believed, arrests were made as a result.

Today, as the web is being undermined by the rapid dominance of apps for smartphones and tablets, the Iranian police would probably, as the jargon has it, ‘go multiplatform’. That, at any rate, is what their colleagues in the Metropolitan Police have just done: unveiling, ahead of the Olympics, a new app called Facewatch ID.

You enter your postcode, and are shown CCTV images of people committing minor crimes in your area. If you recognise anyone, you can let the police know straightaway using the app. The idea is that, if popular, the app will work as a deterrent while also allowing the general public to share in the burdens – and joys – of policing.

The police – or the people who provide them with the technology – may be hoping for other, less obvious benefits from the new system, which could help prepare the ground for fully automated facial recognition systems. The old inefficient approach – putting up wanted posters at bus stops; policemen showing photos of suspects to potential witnesses – gradually gives way to a hybrid approach, where the photos are distributed via smartphones, only to be replaced by a fully automated approach, where human intervention becomes unnecessary altogether. This isn’t to say that automation gives perfect results – far from it – but it’s easier for its advocates to make the case for complete automation once a hybrid system is already in place.