Megatrends and How to Survive Them #11 The Next Technology Revolution
Megatrends and How to Survive Them is the title of our book that is published by Cambridge Scholars Publishing. This is one of a series of blogs based on the work we have done for the book. This blog is about The Next Technology Revolution. This revolution covers advances in robotics, Artificial Intelligence (A.I.) and machine learning, advanced materials, biotechnology and genomics. Biotechnology and genomics are covered in the next blog in this series. We chose to focus on A.I. below because of the wide impact it will have across all organisations within our timeframe.
A.I. is embedded in many systems to solve business problems, from robotics and autonomous vehicles to 3D/4D printing and machine learning. It links into Big Data (the analysis of Big Data uses A.I.).
This trend has strong links with Connected World, of course, and it impacts Social Structure and Economic Activity as one of the drivers of change in society and the economy. Shifts in population – ages and densities – will have impacts on what kinds of A.I. need to be developed. Migration and Urbanisation will lead to smart cities with needs for different kinds of A.I. as well. Both of these will impact what is developed and likely, in what order.
Through the development of A.I. what was once science fiction is fast becoming reality. The Qualcomm sponsored Tricorder XPRIZE was won last year (see the link below). A.I. will transform all aspects of our lives, from wars, crime and justice to our jobs, society and social structures. It will transform what it means to be human. A.I. is seen as “the new oil” because of its potential contribution to the world economy.
The evidence today is that the centre of gravity for A.I. research is shifting away from North America to China. At a recent annual meeting of the Association for the Advancement of A.I., which has been setting the standard since 1980, China submitted more papers than the USA and only three less were accepted. The figure tracks the patents in A.I. across major patent agencies, as shown in the Figure below.
The use of A.I. in decision making – whether in embedded systems or in explicit decision-making tools for managers or front-line workers – can raise concerns in the global North about potential biases and assumptions which are built in, intentionally or not. Machine learning is very exposed to algorithmic bias, where the data used can reflect the implicit values (and unconscious biases) of the humans involved in the data collection, selection or use. It has been seen to have an impact on search engine results, social media platforms, privacy and racial profiling (see link on ethics below).
As A.I. becomes smarter than we are and becomes autonomous, it may not make the same decisions that we would, and we may not be able to figure out why it made the decision it did. AlphaGo made a move no one could explain while beating the world champion. What might the implications be for this?
There are many areas of A.I. that will touch us by 2032. Some examples are:
- Many jobs will use A.I. assistance in some form. Already it is being used in surgery and in law.
- Pharmacies (and warehouses) can use robots to pick and pack medicines or stock.
- Cleaning and personal care is likely to be widely undertaken by robots and there is already a British Standard covering their use. As populations in some countries fall, there is likely to be an unfulfilled need for care which robots can take over. This is already beginning in Japan.
Just as in any product or services, those that incorporate A.I. will go wrong on occasion, requiring specialist diagnostic tools and staff trained to do the diagnosis.
The impact on consumers and providers is likely to be high. As purchasing decisions are delegated to A.I. (e.g. food ordering from smart refrigerators, algorithms predicting which brands you will want to buy etc.), the way that companies reach their customers and engage with them will be transformed. Brands will be compelled to compete against each other electronically, but even within this space, these systems could reflect developers’ biases or completely block out brands that actually meet the consumer’s criteria. It will make it very difficult for companies to engage directly with their customers. Smart assistants and connected devices could even be influenced by sponsors who fund the applications, either openly (for the consumer) or not.
These are just some of the issues that may arise, and, in this blog we haven’t looked at the ethics and decisions that will need to be made before we get this far; you will find more about that in the book.
Some questions that might be useful for you to explore:
- How might A.I. disrupt your business model?
- How will A.I. affect the skills of the people you will need?
- How will your customers relate to you if all services are automated?
- How will you deal with product or service failure due to A.I.?
- How might products and services supplied on a subscription model present new quality and safety challenges, as costs of failure will fall more on the supplier?
Some interesting articles in this area:
Building a real-life tricorder, a la Star Trek:
1/3 of Americans would prefer a robot to a human boss:
How ethics needs to come before the design and development of A.I.
With thanks to David Smith for his thoughts on a number of these aspects of A.I. – we live in interesting times!
Authors: Patricia Lustig, MD LASA Insight and Gill Ringland, SAMI Emeritus Fellow and Director, Ethical Reading.