Rise of the machines

OK, most of us know by now, or at least we should know, that lax use of technology coupled with the internet can leave our very personal information open to the world. Yet, with never-ending technological advances, even the most cautious among us may be exposed to hackers. You could put it all down to the rise of the algorithm; the 'brain code' for computers.

OK, most of us know by now, or at least we should know, that lax use of technology coupled with the internet can leave our very personal information open to the world. Yet, with never-ending technological advances, even the most cautious among us may be exposed to hackers. You could put it all down to the rise of the algorithm; the 'brain code' for computers.

In other words, computers, and the software they run on, are becoming smarter. As is usually the case with new developments, there is a good and a bad side. More intelligent machines can, and do, benefit mankind and the world in general. Or even if they don't, they can be just harmless fun. Take Microsoft's recently launched how-old.net, which uses advanced algorithms to guess the age of someone in a picture. Similar technology is used in Google's image-searching tools and Facebook's tagging feature.

The more sinister elements, however, are worrying. For example, an algorithm was created that allowed users to bypass the privacy settings of Photobucket to locate and copy nude photos on private accounts. Thankfully, the creators were arrested by the US Department of Justice. Other problem areas are when potentially sensitive information is gathered in government surveillance and for advertisement revenue.

Another snag with 'thinking machines' is that they learn from user interaction. Basically, 'smart' software analyses the information it gathers from your technology use to create what you might call your 'personal profile'. But is it always accurate? The evidence so far suggests no.

Finally, a big issue that still troubles learning machines are human biases that affect them. The fact is algorithms are designed to follow step-by-step instructions to learn about whatever their creators want them to. In this way, flaws are virtually unavoidable. The best advice, it seems, is to take care with the type of information you share. The machines are rising, with thanks to algorithms, but we can keep them at arm’s length.

Prepared by: Jonathan Zur

Top
We use cookies to improve our website. By continuing to use this website, you are giving consent to cookies being used. More details…