How is AI influencing the public sector?
- New potential applications for artificial intelligence (AI) are emerging all the time
- Many of these applications could bring benefits for public sector organisations
- We look at some of the ways AI is being applied in 2018 and the potential risks and rewards
Across the world, countries are looking at how to make the most of the opportunities afforded by artificial intelligence (AI).
In January, Theresa May announced the UK would be joining the World Economic Forum’s council on artificial intelligence, as well as forming an advisory body to coordinate progress on AI with other countries.
But how is AI actually re-shaping day-to-day life in 2018? Here, we look at some of its latest applications, and discuss the potential benefits as well as concerns for the public sector.
Health and social care
Detecting and preventing illness and injury is an area where AI could have a significant impact.
Many developments in this area relate to wearable technology and other smart devices – from Apple Watches that can detect hypertension to smart toothbrushes that can flag up potential symptoms of heart disease.
Machine learning algorithms are also being created that can detect variations in the human voice, which it is hoped could lead to earlier diagnosis of dementia and other neurological or cognitive diseases.
New potential applications for AI in health and social care are constantly emerging. For example, recent reports have suggested virtual digital helpers could have a role in combating loneliness among vulnerable elderly people.
AI is already beginning to have an impact in the education sector, with some higher education institutions trialling robotic virtual assistants. There was also controversy late in 2017 when it was revealed machine learning algorithms could be used to help Ofsted prioritise which schools to inspect.
Some experts have also questioned whether the nature of teaching itself – and the type of subjects being taught – may also need to change to ensure the workforce of tomorrow is equipped with the right skills to thrive in an age of automation.
Information and advice
Artificial intelligence is altering the way people are able to access information and advice about public services.
Virtual assistants – or chat bots – are increasingly being used across the public sector to filter and respond to queries, in order to free up staff time. The NHS, for example, has trialled a system in parts of north London where chat bots can take details of patients’ conditions and then advise them on the most appropriate course of action – from visiting their GP or a pharmacy, to going straight to A&E.
Transport for London (TfL) launched its first chat bot last year, which can interact with people using the Facebook Messenger app. The idea is to make it quicker and easier to access real-time information about transport services. For example, a Messenger user chatting to their friends can receive an update on when their bus is due to arrive simply by sharing their location, rather than logging onto another app or website.
In local government, Enfield Council is using AI to increase the number of resident queries it can handle. It has worked with a US AI company to develop Amelia, a robot that uses natural language processing to interpret emotions expressed in the human voice, so it can respond appropriately to queries from the public.
It is currently being used to deal with planning permission enquiries, and the software developers say it handled 2,300 queries during a three-month period in 2017.
Risks associated with artificial intelligence
While there are many advantages, there is also the potential for increasing risks that may be associated with AI.
Tilden Watson, Head of Education, Zurich Municipal, says: “AI was identified as the biggest trend that will have implications for public services from our own customer surveys.
“Whilst we have identified many positives the concern remains that AI coupled with robotics could produce errors on a massive scale.”
As the use of AI becomes more commonplace, here are some of the questions you may wish to consider.
- Who will monitor the quality of the information your virtual assistants give out?
- Could you be liable for bad advice given by a chat bot?
- How will you ensure the data your AI systems feed on is accurate, up-to-date and free from bias?
- Will reduced human interaction in your processes mean reduced oversight?
- Is there a danger you will become over-reliant on AI, and lose the personal touch when dealing with customers?
- What about transparency? Will the public understand and accept decisions reached by a machine learning program?