Dr Luc Julia, pictured, is vice president of innovation and CTO at Samsung Electronics Samsung Electronics

Dr Luc Julia, pictured, is vice president of innovation and CTO at Samsung Electronics Samsung Electronics

Hollywood films delight in depictions of dystopian futures filled with autonomous computer systems or killer robots, stoking general public distrust and misunderstanding around AI. However, according to Julia, these futures will never come to be. “The artificial general intelligence that we hear about in the media doesn’t exist,” he explains. “AI can do well in specific things, like chess or driving a car, but we are in control, we decide that.” His upcoming keynote in the AI/Digital Summit at ICT Spring will expand on this, covering the history of the technology and the exciting innovations that can realistically be developed around it.

Julia, who held the roles of chief technologist at Hewlett-Packard followed by director of Siri at Apple before joining Samsung in 2012, has recently released a book in French under the title L’Intelligence artificielle n’existe pas. It delves deeper into the possibilities and constraints of artificial intelligence - a technology whose very name he considers misleading. “An intelligence that is above our human intelligence will not exist,” he says. “The human brain isn’t a machine that could be built by us – there’s no way we can even get close to what it can do.”

He explains that what we call AI is based on algorithms developed over 60 years ago, which haven’t changed much since then.
The real breakthroughs in its application came thanks to the internet and the availability of big data. “AI was able to use machine learning and statistical algorithms to look at massive amounts of data and make real-time decisions,” he explains. This led to the most well-known examples of machine “intelligence”, when a computer first achieved the feat of beating the top player in the world at chess in 1996, followed by the even more impressive win by a computer against the reigning Go champion in 2017.

Julia explains that the most significant developments in AI in recent years have been in image recognition, especially applications in the medical field, and in speech recognition, exemplified by virtual assistants like Siri and Alexa. He sees the potential for widespread progress coming to the medical field, combining the power of AI with big data to better detect diseases using imaging. A proliferation of assistants will also appear to help us with everyday tasks, controlled not only by voice but also the power of the body. He also predicts huge improvements in the transportation industry, not due to what he considers the fictional idea of fully autonomous cars, but by using AI to make cars much more secure to prevent countless accidents.

Julia sees no inherent danger in AI technology, despite the widespread myths in popular culture. Rather, he is more worried about the stagnation of progress that could be a result of them, warning that the life-changing developments he predicts will only be possible if we can better educate people on what AI can and cannot do on its own. His mission to dispel these myths is fuelled by his positive vision of the future and the fear that it could be held back by unfounded beliefs. “I want the technology to develop new things as there are a lot of good things to be done. But when people are too afraid, they just want to stop the advances. I don’t want a new winter of AI,” he says.

This article was first published in May 2019 the ICT Spring 2019 supplement of Paperjam magazine.