AI Will Become Ubiquitous in Five Years' TimeAI Will Become Ubiquitous in Five Years' Time
AI Will Become Ubiquitous in Five Years' Time
April 3, 2017
We spoke to The University of South Wales' Senior Lecturer in Computing, Bertie Müller, about how AI is currently being used by businesses across a wide variety of sectors, and how the new technology will affect industry moving forward.
Berndt Müller, referred to as Bertie, is the Senior Lecturer in Computing at the University of South Wales, and he'll be a keynote speaker at The AI Summit which is taking place in London on May 9th and 10th. From September 2006 to September 2010, Müller worked in the Department of Computer Science (later known as the School of Engineering and Computing Sciences) at Durham University, and is also Chairman of the AISB (Society for the Study of Artificial Intelligence and Simulation of Behaviour) and Associate Editor of the journal Connection Science. In short, Mr. Müller is a wealth knowledge on all things AI-related.
We started off our conversation with Müller by asking him to elaborate on the areas of business he sees AI having the biggest effect. "An immediate effect can be seen in the financial sector," he began. "Banks and insurance companies already make many decisions based on big data analytics. The next wave of big effects will be seen in customer services across sectors, with automated text-based chat support leading the way."
Bertie then stated that healthcare will undergo changes thanks to AI. "Another area that will see some changes is healthcare, though regulations will mean a slower adoption of AI techniques than in other sectors."
Being an academic, we were keen to find out how academia and business might collaborate in relation to AI. "Academic research in AI has for many decades been viewed as a hobbyhorse and as pure utopia. The recent resurgence of AI has been triggered by the availability of powerful hardware and shows how important early research in AI has been," replied Müller.
"At this stage it is more important than ever that academics are respected in their role as innovators as well as advisors with a responsibility to educate decision makers in industry as well as the general public about possibilities and risks of AI. Many AI techniques work reliably under certain constraints only. These constraints can be extremely complex and ignoring them can have disastrous effects. It is appreciated the not all companies will have the resources to employ their own R&D team or AI experts, and even if they do, it should become common practice to seek the advice of academia," he revealed.
"We are seeing a surge in partnerships between SMEs and academia in various areas related to AI to give them an edge over their competition and establish niche markets in which they can thrive. Academia should inform all sectors at all stages of AI facilitation," he finished.
AI is one of the buzzwords of 2017. It seems as if this new technology has taken the business world by storm, so we wanted to know what Müller thought was going to be the rate of adoption of AI industry wide this year, and how it'll change things for business in the future. "This is a hard question," he admitted [sorry about that Bertie].
"AI is often hidden under the bonnet and it depends on what you count as AI adoption. It is quite clear that the use of AI will have a tangible effect in almost all sectors from logistics to healthcare. Vertical markets will be the forerunner, to be followed shortly by AI-based solutions in horizontal markets," explained Müller.
Yet, it's not as if companies wanting to adopt AI will be able to do so easily. There will be challenges, and we asked Müller to elaborate on what he believed would be the biggest challenges for businesses looking to adopt AI, and how he thought they would tackle them. "On the technical side, one of the main challenges is the (non-)transferability of methods across different domains. It worries me how carelessly some companies use methods that are guaranteed to produce meaningful results under certain constraints even when these constraints are not met," revealed Müller.
"A single success cannot justify widespread use without proof of the method's accuracy in the newly adopted scenario. All application of AI should be under the scrutiny of an ethics panel ensuring basic moral and ethical values," stated Müller.
"Companies need to make sure their customers/clients (and their employees) feel informed about and are comfortable with the use of AI. Uncertainty leads to distrust and this cannot be to the benefit of any company, so forthright information about the use of AI and disclosure of the nature of a conversation partner in electronic communications or the grounds of a decision can help establish a relationship of trust," expressed Müller.
Since advancements in AI technology have been happening at such an impressive rate, we wanted to know where Müller thought the adoption of AI would be in five years' time. His answer was short, but sweet. "It will be ubiquitous, but it will still be specialised AI solutions rather than a general AI," he finished, and with that, our time with the University of South Wales' Senior Lecturer in Computing had come to an end.
You can hear more of Bertie Müller's thoughts on AI in business during his keynote presentation at The AI Summit in London, which is fast-approaching (May 9th and 10th).