Mon | Dec 17, 2018

Artificial Intelligence is a great tool - IBM director

Published:Tuesday | December 4, 2018 | 12:00 AMMark Titus/Gleaner Writer
Nicholas Fuller, director for Global Research Innovation at IBM Technology Services

Despite fears that rapid automation and digitalisation could reduce the number of jobs in the manufacturing sector, Nicholas Fuller, director for global research innovation at IBM Technology Services, believes the imminent role of artificial intelligence (AI) on the global landscape is more an enabler than a threat and must be embraced.

"I don't view it as a cause for concern; I see it as an opportunity," said Fuller, whose role at IBM includes building and incubating technology for his company's business unit and leveraging technology such as AI, cloud and augmented reality.

"If you have different skilled workers working on a problem in a call centre, for example, what artificial intelligence does is to help those agents do their jobs better, become more efficient."

He continued, "Those agents, who I expect should be in the majority, who embrace the tooling, will have nothing to fear because it helps them do their jobs better. It is a huge opportunity for them to improve on their current jobs and essentially help their firm deliver value in a more efficient way. However, those who are not willing to undergo the cultural and educational transformation required to use the tooling could be affected."

 

AI IS ABOUT AN ALGORITHM

 

Fuller, who is described as a master inventor on his company's profile, acknowledged that there are currently disadvantages with AI, also called machine learning, including the fact that it is only as good as the training data it receives, but says that his firm could have the solution.

"AI is about an algorithm that is able to do some things for you that you could not do before," he told The Gleaner following his presentation at Tech Beach Retreat at Iberostar Hotel in Montego Bay, St James, on the weekend.

"Whether it is understanding the text, whether it is conversing with you to troubleshoot a problem, if you don't have a training data that represents the full range of possibilities in terms of outcomes, then it will perform with an inherent bias, and we don't want bias in our computer algorithms; that's a problem."

... Another five to - 10 years for AI to dominate globally

IBM recently launched a programme called AI Overscale, which is meant to detect bias and address it in a machine-learning algorithm, but Nicholas Fuller, director for global research innovation at IBM Technology Services, is of the view that it will take another five to 10 years for artificial intelligence (AI) to dominate globally.

"We use AI today to understand the unstructured text, for conversation and chat, but that is the beginning," said Fuller, who hails from Trinidad and Tobago.

"When you think of a problem in healthcare, of trying to diagnose a heart issue of cardiac MRI, that is a complex problem. To train a model accurately is a difficult problem. When you think of the support domain of being able to provide predictive application failure for a large enterprise such as a bank, being able to ensure that that works well during a period of stress, like Black Friday with real-time transactions happening, those are tough problems, and we are just at the beginning of that, so domination on a global scale in a developed concept is a good five to 10 years away."