From intelligence amplification to artificial intelligence
From self-driving cars to self-learning robots, artificial intelligence applications are now hitting the mainstream. But according to Fujitsu’s CTO for EMEIA, Dr Joseph Reger, the true potential of AI has yet to be realized.
Artificial intelligence (AI) has been a widely used — and often abused — term in the IT industry for more than four decades. But over the past five years, such machine-learning technologies have evolved at a pace that has surprised even the closest observers, with signs that genuine AI is beginning to emerge.
Champions of AI point to the recent victory by DeepMind’s AlphaGo computer over the world’s top Go player — the profoundly complex board game — or the incorporation of AI in industrial robots that learn as they age. And for Dr Joseph Reger, a Fujitsu Fellow and the global ICT company’s CTO for Europe, Middle East, India and Africa, such developments just underscore the accelerated transition to truly smart machines.
“What is commonly referred to as artificial intelligence is moving into the business very visibly. Just in the last year or so, there have been enormous advances,” says Reger in I-CIO’s exclusive video interview. Inspiring and driving that is a desire by business and government organizations to embrace the digital revolution: “A large number of businesses seeking disruptive innovation and digital transformation are now using some sort of self-learning, deep-learning intelligent algorithms,” Reger says — not only to reinvent their products and business models but to change the nature of their industries.
Reger, however, characterizes this as part of a journey — albeit a fast-accelerating one. “Is what we’re seeing today real artificial intelligence? I don’t think that most of the projects that the technology industry does currently are truly AI.”
And he elaborates: “Indeed, anyone would have a hard time to explain the difference between AI as it stands today and IA — intelligence amplification. IA is like a mighty exo-skeleton for the brain. It assists human intelligence to make better decisions, to work faster and so on.” And although the algorithms being applied have become much more sophisticated, he maintains that “most of what the technology industry does currently under the label of AI is actually intelligence amplification.”
Machines that learn
What is remarkable is how quickly the AI-enabling technology of neural networks has matured, and how machine learning is becoming part of the business landscape as a result of that. “Learning algorithms are here, and we know how to use deep-learning networks to provide systems that react to unknown situations, that make decisions in situations that were never foreseen by their creators.”
Over the next five years Reger believes the impact of such activities will be increasingly visible in many areas of industry and society — from self-driving cars to domestic service robots. As for the term artificial intelligence: it will continue to be liberally — and loosely — applied, he says.
• Dr Joseph Reger will be delivering a keynote address on artificial intelligence at Fujitsu Fourm 2016 in Munich, 16 and 17 November
- Photography: Enno Kapitza