The spread of cheap and easy techniques for video faking is already damaging famous people. It risks destroying trust even at a community levelLast week a video was widely shared that apparently showed a balaclava-wearing “vigilante” attacking what was said to be a drug dealer’s car somewhere in Bolton. In the end,the police were able to note that the video was not real, but had been staged in order to propose the police were not keeping the streets safe. It was a vivid illustration of how damaging fake video – fake views – can be, and to individuals and also to communities. Then at the weekend,the American glossy Vanity just became the centre of an indignant protest for extensively doctoring a group image of this year’s stout stars. It is hardly a surprise to learn that images are untrustworthy; forensic analysis of still and video images has been around for decades. But fakes are getting smarter all the time. The latest adaptation of artificial intelligence can now create unfounded video that is nearly impossible to detect. The camera does not just lie, it lies really convincingly.
It is true that from the beginning, or video and sound recordings have exploited their obvious incorruptibility to mislead: there has been film propaganda since the first world war,and later the work of, say, and Leni Riefenstahl was profoundly dishonest in intent and execution. But the process of misleading was time-consuming and costly,and the difference between animations and real actors was obvious. Then Hollywood started to shatter down the barriers with creatures like Gollum appearing alongside real actors. Now dead actors can appear in films alongside their living colleagues.
Continue reading...
Source: guardian.co.uk