The Future of Data Use

Two completely different articles have today given me pause to think the future of data analytics and machine learning, the vision for the future and the steps we, as Australians, need to take to grasp the opportunity and remain relevant. The first article discusses the future of the Google Assistant and mechanisms Google used to collect good quality voice data to train voice recognition algorithms. 

The mechanism for doing this was the creation of a telephone service, 1–800-GOOG-411, to offer free number lookup and connection.  I am struck by the vision of a company that will invest time and resources to create service to collect voice data to develop a machine learning model that better can understand the nuances of speech. Natural Language Processing is interesting in its own right, but more especially so when you look at it as a foundational step towards digital assistants.

The article here is well worth a read for anyone interested in the future of digital assistants. Do we in Australia have startups, established businesses or government agencies with the same level of commitment and vision for driving value from data for the purpose of innovation?

The second is the recent draft Productivity Commission Report on Data Availability and Use which although well considered and thoughtful feels like a conversation that should have been bedded down 4 years ago. Don’t get me wrong I agree with much of the report – yes we are missing opportunities to derive better outcomes for Australians, yes there are risks that need to be considered, yes the opportunities need to be balanced with considerations for the rights of individuals.  It also raises the important question of who owns data – does data generated by my vehicle when I am driving belong to me, my insurer or the car manufacturer. Each party has a vested interest in having access to this data.

One of the interesting observations is the sheer amount of information that Australians already give away via social media apps, wearable devices and so-called free applications that harvest data for a variety of purposes.  Users assume that Pandora, Netflix and Google use data to generate recommendations and they are comfortable with that, given that the suggestions back add value and often don’t come with an overtly commercial angle. However when it is the government, bank or local supermarket doing the same it can feel invasive or unwelcome. 

Of significance is the insight that many Australians are quite prepared to share de-identified heath data to advance medical research and improve patient care – again potentially as there is a level of self-interest and philanthropy. The cited examples include New Zealand use of longitudinal data to identify youths at risk of poor outcomes in adulthood. Similarly the UK’s use of administrative hospital records linked (via unique patient health service number) with a number of cancer screening registries to improve how and when cancer is diagnosed. There is already strong support for using Australia’s health data in research. A recent survey revealed that over 90 percent of Australians were willing to share their de-identified health data to advance medical research and improve patient care. 

The opportunities for health and patient outcomes extend far beyond this. In fact I believe that the greatest opportunities for Big Data Analytics are those that focus on health - by digitising and effectively using health data potential benefits include predicting and detecting disease earlier, managing patient care at an individual level, identifying patients at risk from any number of complications, enabling real time analytics and alerting of life threatening conditions, and as input into informed strategic planning. 

In addition, if we fail to create an effective means to share, analyse and derive better outcomes from data we are doing a disservice to both patients and limiting our economic opportunities.

Whether we like it or not data analytics and machine learning will form a part of the future in which we, as a society, operate. There are quite profound implications for individuals, government agencies and businesses. The importance of the Productivity Commission Report is that the “data use” conversation has been kick started, the opportunities identified and the data privacy challenges, as a result, will be dealt with. 

Of greater importance is the recognition that this is just the first step – data analytics, machine learning and artificial intelligence are the foundation for a new suite of technology solutions, such as Digital Assistants and Driverless Cars, which will have a huge impact on our society, on the way we work and the way we live.