Last year, Scottish Local Government Chief Digital Officer Martyn Wallace spoke to the CIO UK podcast and highlighted that in 2019, local government must take advantage of artificial intelligence (AI) to deliver better outcomes for citizens. He explained:
“I think in the public sector we have to see AI as a way to deliver better outcomes and what I mean by that is giving the bots the grunt work – as one co-worker called it, 'shuffling spreadsheets' – and then we can release staff to do the more complex, human-touch things.”
To date, few councils have had the capacity to invest in AI. However, the mood is slowly starting to change and there are several examples in the UK and abroad that show artificial intelligence is not just a buzzword, but a genuine enabler of change.
In December 2018, Local Government Minister Rishi Sunak announced the first round of winners from a £7.5 million digital innovation fund. The 16 winning projects, from 57 councils working in collaborative teams, were awarded grants of up to £100,000 to explore the use of a variety of digital technologies, from Amazon Alexa-style virtual assistants to support people living in care, to the use of data analytics to improve education plans for children with special needs.
These projects are still in their infancy, but there are councils who are further along with artificial intelligence, and have already learned lessons and had measurable successes.
Milton Keynes Council has developed a virtual assistant (or chatbot) to help respond to planning-related queries. The AI project – which integrates with the back office infrastructure provided by Idox – will deliver three key outputs:
With Planning a highly-skilled profession, the chatbot has had to be able to deal with technical questions and complex scenarios. Trials have shown that the virtual assistant is better able to validate major applications, as these are often based on industry standards, rather than household applications, which tend to be more wide-ranging.
Chief planner, Brett Leahy, suggests that introducing AI will help planners focus more on substantive planning issues, such as community engagement and placemaking, and let AI ‘take care of the constant flow of queries and questions’. Following the chatbot’s launch, the volume of calls to the council’s Planning team has already seen a reduction, while the number of conversations with the chatbot is increasing, saving important planning officer time.
In Hackney, the local council has been using AI to identify families that might benefit from additional support. The ‘Early Help Predictive System’ analyses data related to (among others) debt, domestic violence, anti-social behaviour and school attendance, to build a profile of need for families. By taking this approach, the council believes they can intervene early and prevent the need for high-cost support services. Steve Liddicott, head of service for children and young people at Hackney council, reports that the new system is identifying 10 or 20 families a month that might be of future concern. As a result, early intervention measures have already been introduced.
In the US, the University of Chicago’s initiative ‘Data Science for Social Good’ has been using machine learning (a form of AI) to help a variety of social-purpose organisations. This has included helping the City of Rotterdam to understand their rooftop usage – a key step in their goal to address challenges with water storage, green spaces and energy generation. In addition, they’ve also helped the City of Memphis to map properties in need of repair, enabling the city to create more effective economic development initiatives.
Yet, like most new technologies, there has been some resistance to AI. In December 2017, plans by Ofsted to use machine learning tools to identify poorly performing schools were heavily criticised by the National Association of Head Teachers. In their view, Ofsted should move away from a data-led approach to inspection and argued that it was important that the ‘whole process is transparent and that schools can understand and learn from any assessment’.
Further, media reports have led to a general unease that introducing AI could lead to a reduction in the workforce. For example, PwC’s 2018 ‘UK Economic Outlook’ suggests that 18% of public administration jobs could be lost over the next two decades. Although its likely many jobs will be automated, no one really knows how the job market will respond to greater AI, and whether the creation of new jobs will outnumber those lost.
In the next few years, it’s important that local government not only considers the clear benefits of AI, but also addresses the public concerns. Many citizens will be in favour of seeing their taxes go further and improvements in local services – but not if this infringes on their privacy or reduces transparency.
Pilot projects, therefore, which provide the opportunity to test the latest technologies, work through common concerns, and raise awareness among the public, are the best starting point for local councils looking to move forward with this potentially transformative technology. Once tried and tested, the question and mystery surrounding the authenticity and true value of AI may finally be answered.