I encourage physicians and healthcare executives to read the 2017 Health and Technology journal article: Google DeepMind and healthcare in an age of algorithms, the cautionary tale of the behind the scenes (and somewhat questionable) relationship between part of Great Britain’s National Health Service and Google.
On November 18, 2015, Royal Free London NHS Foundation Trust started the transfer of millions of patient records to DeepMind, an Alphabet (i.e. Google) subsidiary.
To make a long story short, Royal Free – without any significant patient, citizen or government oversight- elected to transfer vast patient data sets to an external Artificial intelligence (AI) company, DeepMind, with very little legal restriction on how this data can be used.
This paper brings up significant ethical questions, including topics regarding the extent of patient rights of privacy and informed consent; these make the paper worth reading alone.
However, and perhaps more interesting, it points out the need to consider large population data sets as a public commodity, with an intrinsic value, and us, as the generators of that data as we go about our lives, stakeholders in that data.
For example, most Americans would be opposed to giving our National Parks to a realtor secretly, without debate, so they can develop the land for their own economic interests, – no matter how much the realtor said it would be for the public good.
Then why, then, is it acceptable for the British NHS to give Google – effectively the large advertising company in the world – data on millions of people for years without significant public scrutiny and discussion?
Simply put. It is not.
Look – the value and quality of artificial intelligence output is dependent on the quantity of the data. No data, then no AI.
As an individual with individual data, we have no leverage over the Googles, Apples, Facebooks, Microsofts, Amazons and IBMs of the world.
It is a one way mirror. They can see and know everything about us, but we have little knowledge of their internal workings.
But as a citizenry we could have leverage, – if, and only if, we value ourselves and our group data.
Anyway, as physicians, we need to be patent advocates not just in the real world but the digital one, and the first step towards this is understanding how data is being used in AI.
Reading the above linked article is a good first step.
Enjoy.