As you may have guessed, this iPad 2 user was the technology entrepreneur I had come to meet: Mike Lynch, co-founder and chief executive of Autonomy, the British software house. I wanted to hear about the augmented reality app his firm has just developed – in part because I couldn’t fathom the link between AR and the firm’s claim to fame to date: predictive software.
Autonomy has gone from nothing in 1996 to a firm worth £7 billion today by leveraging the theories of an 18th-century English mathematician and cleric called the Reverend Thomas Bayes, who worked out how to calculate the probability that certain variables are associated, whether they are words, behaviours or images. Lynch and his colleagues built their business on a pattern recognition engine called the Intelligent Data Operating Layer (Idol) that uses algorithms based on Bayes’ ideas.
On the London tube, Autonomy’s Bayesian algorithms that analyse CCTV images to calculate the likelihood someone will try to commit suicide by jumping under a train – allowing the track current to be turned off and help sought. If you follow Formula 1, Ross Brawn’s Mercedes F1 team identifies the potential source of every advantage gained by rival teams by training Autonomy algorithms on post-race video. To prevent fraud or noncompliance with the financial laws, workplace emails are analysed to infer risk. And police can use Idol to seek hidden patterns in crime reports.
But that’s in the PC world. Now, says Lynch, they want to exploit the awesome and growing power of smartphones like Android and iPhone/iPad. To do this they have written an app they’ve called Aurasma that allows anybody to associate real world items with online content, which they liken to an aura – hence the name.