Reimagining government: the power of AI

Reimagining government: the power of AI

This is the first in a series of articles looking at how governments can employ emerging technologies to transform the way their citizens live, work and play. Using case studies and examples, and drawing on my own experiences and EY’s expertise, I aim to show how innovation can help governments deliver everything from improved value for money to smarter infrastructure, as they seek to build a world that works better for citizens. By leveraging new technologies and forging new public-private partnerships, governments will solve problems and create smarter societies, helping to build nations, regions and cities that can meet the challenges of the 21st century.

In each article, I will look at one emerging technology. The first is Artificial Intelligence (AI).

You’re about to start the journey to work when a phone alert tells you that trains are delayed because a tree has blown down on the line. Your phone presents you with a variety of choices. Cycling, it says, would be best in terms of your health. However there is an 85 percent chance of rain, so it has pre-ordered you a self-driving car because it knows you don’t like to cycle when it’s wet. Besides, you can use the time in the car to prepare for your first meeting of the day. Even with rail disruption, AI transportation management and self-driving cars mean that traffic is flowing smoothly and your journey will only take two minutes longer than usual.

Most people imagine AI as robots and computers that talk to us as equals. This version of AI (which is sometimes called “Strong AI”) means conscious machines that think as we do and is almost certainly still decades away. But “Weak AI” – which entails highly sophisticated machines that give the impression of intelligence, perform some tasks better than humans, and learn and adapt from data they are given – is here already.

Chatbots are an obvious example. They appear intelligent and are capable of learning within limits based on the conversations they have. In my discussions with government in India, we spoke about adopting algorithms and self-driving cars to enable the future of mobility. Another example is systems that grade papers at some US universities.

AI has already shown itself to be superior to humans in some situations. A recent study of emergency calls undertaken by the University of Copenhagen showed that software produced by the Danish AI company Corti SA correctly detected cardiac arrests in 93 percent of cases, compared with 73 percent for humans. What is more, with AI the diagnosis was made more quickly.

We are going to see a lot more of this sort of thing in the public sphere. Sometimes the applications will be almost invisible – in areas such as traffic management and waste removal. Sometimes they will involve the replacement of frontline staff – such as in healthcare and the issuing of permits.

In cities, AI will also involve a lot of smart systems working together. If, as in our future scenario, a railway line is blocked, AI will be able to not only reroute trains but also adjust other networks such as buses and traffic management to take the problem into account. It won’t be seamless at first. But slowly, and then much more quickly, the world will start to look like our commute of the future.

Other infrastructure possibilities range from power grids (including home generation) to water supply, from hiring rooms in smart buildings to waste bins that know when they need emptying and tell a system that plans the most efficient way to empty them. All this will, unsurprisingly, involve the internet of things and the growth of smart infrastructure. The consultancy Gartner predicts that 11.2 billion things will be connected by the end of this year and 20.4 billion connected by the end of 2020.

There are numerous other applications too, many of which are less obvious. One, which I’ve discussed with government in Dubai, is detecting fraud in welfare systems (AI is very good at spotting anomalies in systems). Another is the prediction of events such as flooding, and the implementation of solutions and defensive measures before the flood occurs. Others include working out the total return on investment from difficult to quantify schemes (such as bike hire or reduced price transport), modelling human behaviour in cities and more efficient taxation.

Finally, AI may be the answer to numerous legacy problems that many governments have. One is huge backlogs of applications or cases. Set an AI to do them. Even if it can only deal with the easiest 75 percent of cases, it will still save huge amounts of time and money, and deliver benefits for citizens. AI is already being used in legal casework in the UK.

The other great legacy application is the huge amounts of data that governments hold on their citizens. It is often “unstructured” and scattered around on systems that are incompatible with each other. Until recently, this has meant everything from data being laboriously inputted by hand to missed opportunities in areas from health to finance because systems are siloed. Increasingly, AI can make sense of this data. Moreover, it can also take legacy systems and migrate them to the Cloud.

Cloud storage is cheaper and more efficient. Integrating systems means they can talk to each other – and, especially with the use of AI, deliver new insights that will allow better delivery of services. The data itself will be far more valuable and easy to monetise in its new, structured, integrated state.

The 21st century is a challenging place and the pressure on governments has never been greater. “Do more with less” is a never ending mantra. But I believe in areas as varied as healthcare, transport and criminal justice, artificial intelligence really will allow us to achieve this.

The question is: does government have the imagination and commitment to seize the opportunities that artificial intelligence offers?

Aasma Pratap Singh, MBA

Product Ops leader driving cross-functional alignment, enabling 20+ successful product launches and contributing to $30M in new revenue, scaling to $190M ARR.

5 年

Hey Rohan! Very interesting read! I am working on applying machine learning and artificial intelligence in education management space. Mostly geared at 2 levels: 1. Predictive analysis for preemptive state action. 2. AI to aid mastery learning of concepts.

要查看或添加评论,请登录

社区洞察

其他会员也浏览了