Solving Old Problems with New AI
- By Winston Thomas
- September 10, 2018
AI is not about the future of IT; it is also offering a refreshing approach to solving teething problems that many firms already face.
Speaking at a panel discussion at The Economist's Innovation Summit Asia 2018, entitled Riding the Wave -- The Future Company, Ross McCullough, President, Asia Pacific, UPS noted that AI is helping them to do a lot more and allow them to stay competitive.
McCullough noted that UPS has always embraced technologies, allowing them to deliver 20 million packages and manage 60 million in the system. “Without machine learning, we will not be able to [do this],” McCullough said.
UPS got onto machine learning early. Their logistics prowess relies on ORION, a machine learning algorithm that finds the fastest and the most fuel-efficient way to deliver packages.
McCullough also pointed out another crucial advantage of ORION. It keeps vital knowledge within the company. He noted that in the past hiring and retaining the people with logistics knowledge was challenging, while limited their chances to grow. Orion allowed UPS to grow the company quickly without having to hire and train new talents.
With ORION crunching numbers, it allows UPS to be more transparent about their operations and offer more flexibility to their customers.
“We used to have one service. Now, we can deliver at different times…it is not possible without algorithms and machine learning,” McCullough said.
HSBC is another global company that relies heavily on machine learning. They already use it for Anti Money Laundering, Know Your Client and Profiling processes.
The bank recently launched visual recognition, fingerprint authentication, AI chatbots and Robotic Process Automation, “which is a huge area,” Frank Tong, Global Head of Innovation and Strategic Investments, HSBC said.
The banking giant sees AI and machine learning as an evolution rather than a revolution. It also works closely with technology companies to create solutions around AI, while looking for opportunities to invest in key startups.
Tong sees vast potential in voice analytics for improving customer service and its call centers. While he acknowledges enormous leaps in Putonghua and English, Cantonese recognition is still tricky.
Meanwhile, Steve Monaghan, Chairman, and Chief Executive, GenLife is looking to shape the insurance industry.
“Insurance is an antiquated industry. In a connected world, you can communicate risk, but insurance does not do that. The great thing with machine learning is that it can do this far more efficiently,” Monaghan said.
He is also simplifying the information that insurance forms collect. “If you look at the data points in insurance forms, you only need to digitize a small percentage to remove costs and friction, and offer better service,” Monaghan said.
Once you identify the vital data points, Monaghan's team is looking to link with Internet of Things devices and sensors for better monitoring for risk. "Then you can have a product that pays while you are alive.”
However, all panelists agreed that machine learning needs time to learn. It is also not going to be 100% accurate.
“It is just less wrong than a human. AI is not binary, and there is a learning curve,” Monaghan said.
Winston Thomas
Winston Thomas is the editor-in-chief of CDOTrends. He likes to piece together the weird and wondering tech puzzle for readers and identify groundbreaking business models led by tech while waiting for the singularity.