Ahead of the Chief Data & Analytics Officer Melbourne (4-6 September), we caught up with PwC’s Lead Partner, Analytics, John Studley to discuss their recent Artificial Intelligence Index report and what it means for Australia in particular.
Corinium: John, surprisingly the AI Index shows that healthcare is perhaps the best at exploiting artificial intelligence – why do you think that is? How is the healthcare industry using AI to their advantage?
John: If you look at the areas of early detection, diagnosis, research, treatment options and training, AI machine learning is making an impact in a very positive way. AI is helping patients to stay well and prevent reoccurring symptoms with patients having to go to the doctor less often. Examples include smart belts and fitbits to help monitor overeating and other health related issues. These devices have algorithms and magnetic sensors to help consumers monitor their health levels. AI is also already being used in the detection of cancer. For example, in America, the Cancer Society run approximately 12 million mammograms a year on women, but a high proportion of the those give a false result with one in two women being told they have cancer when they, in fact, they don’t. The use of AI is enabling the healthcare industry to screen mammograms 30 times faster with a 99% success rate in some cases, resulting in a better diagnosis, less biopsies and less stress for patients.
Corinium: In your opinion would you say that the full potential of AI is being realised across all industries?
John: There seems to be higher levels of data availability in healthcare and financial services with these industries more likely to adopt AI more quickly. In financial services, AI is used to provide smart projections from large amounts of data. In a service based industry, AI can assist with personalised offers, ads and chatbots. You see this with Amazon for example where AI is used to predict logistics and sourcing. AI is also used in insurance for cyber, fraud and trust to help organisations get to know their customers and prevent anti-money laundering.
“AI could be working much harder in organisations to solve real business problems, making sure we do something actionable with those insights to create value.”
Corinium: The AI Index forecasts that AI will boost China’s GDP by 26% in 2030 and America’s GDP by 14%. What is your view on Australia’s readiness to adopt AI?
John: I think AI is really coming into its own heralded by large increases in computing power, decreases in storage cost, increases in communication speed and advanced algorithms that can be applied to real world problems. I would recommend Australian organisations to really think about what they are trying to achieve, how to create new value, and then look for that as a tool or accessory to deliver on.
Australia as a nation has not yet fully embraced the power of analytics as much as it could. AI could be working much harder in organisations to solve real business problems, making sure we do something actionable with those insights to create value.
“AI systems need to be taught and they need to learn so you should expect there will be the need for diligence around results to make sure the machines are doing what you expect them to be doing.”
Corinium: What would you say are the fundamental steps in ensuring that your business is ready to solve problems with AI?
John: First of all, be careful with your customers’ data. Organisations are increasingly turning to partners and collaborate with other organisations to manage ecosystems but this also brings with it risk. The second is to have a strategy in place and the third point would be to tailor your technology. Once you move past robotic process automation and start to look into augmentation and then autonomous intelligence there’s not so many off the shelf applications. You will need to do a bit of work to customise and bring things together. And lastly, build in the appropriate governance and controls. AI should be managed with the same discipline that you would use for any other technology project. AI systems need to be taught and they need to learn so you should expect there will be the need for diligence around results to make sure the machines are doing what you expect them to be doing.
You can download PwC’s full AI study report here: www.pwc.com.au/artificial-intelligence-study
John Studley, Lead Partner, Analytics, PwC
You can talk more to John at Chief Data & Analytics Officer Melbourne, taking place on the 5-6 September, where he will be available to discuss your AI and analytics challenges in more detail.
For more event information visit: www.chiefdataanalyticsofficermelbourne.com