We must question everything we see. That is what ‘the pioneer of deepfake technology’ Chris Umé told VRT in the course of the broadcaster’s VRT Check Weeks. “The technology is here today. It is very difficult to know what is real in the world we live in now.” Umé is also keen to emphasise the benefits of AI for the medical sector.
Fleming Chris Umé, world-famous for deepfakes, says: ‘We must question everything we see’
Colin Clapson
Published:
Worldwide Chris Umé is regarded as the pioneer of deepfake technology, ever since he shared his “DeepTomCruise” videos four years ago. ‘Without realising it at the time, we changed the world with those videos,’ he says. Now he is warning of the dangers of AI.
‘I posted those videos to warn people: the technology is coming. And now it’s here.’ And it is not without its dangers, according to Umé. ‘As humanity, we must question everything we see or trace it back to the source. The importance of fact-checkers and journalists has never been greater. It is very difficult to know what is real: reliability is all the more important.’
The technology itself can help prevent abuse. ‘Everything needs to be balanced. The AI power used to create fake videos can also be deployed by journalists for fact-checking.’
‘A journalist could deploy a team of 50 AI agents to verify whether an image is genuine or not. They can perform everything a human can do online. One investigates, another deals with audio, yet another searches for similar items. By the time the journalist has taken just one step to carry out the check manually, the agents are already ready.’
Some systems, such as Adobe and Google, have already built watermarks into their AI systems. How reliable are they? ‘Someone who really wants to will simply remove them. You can extract metadata or a watermark from a video by filming your screen. It’s very important that it’s there, but you can’t assume that everyone has good intentions.’
'AI will diagnose flu'
Umé also highlights the benefits of AI for the medical sector. ‘Technology will be capable of things we simply aren’t yet able to do. In the medical sector, for example, AI will be able to detect that you’re coming down with flu based on your behaviour.’
What timeframe is the expert referring to? ‘The technology is incredibly powerful. I think it’s actually already possible, but now it’s up to us to use it properly and connect the right things.’
Ambitious
‘I want to do something positive for humanity with the technology that’s available,’ says Umé. ‘I’d very much like to collaborate with academic scientists and hospitals. I’m not a doctor, but I know what the technology is capable of and what we can technically achieve with it.’
‘We will be able to cure diseases we could never have imagined. As humans, we simply cannot think that far ahead. That technology will amplify that ability many times over, enabling us to detect issues more quickly and make connections to find solutions.’
‘A good example of this is Elon Musk’s Neuralink. I believe this kind of technology will only make things better.’
Furthermore, Umé does not believe in using AI to replace people. ‘My aim is always to empower people and make them believe they can do something they couldn’t do before. That can mean all sorts of things: getting better at their work, for example, or thinking more creatively.’