Artificial intelligence is everywhere. The discipline, which up until recently was restricted to academic research and applications, is finally becoming part of everyday life. Ask anyone. They might not be able to tell you what it is, but they will know that it’s permeated all sorts of industries.

In education, platforms like Century are helping teachers provide more personalised education programmes and receive feedback. In the legal service industry, AI is helping lawyers to do legal searches and draft the best standard documents.

In travel, companies are developing fully autonomous systems that diagnose potential problems for driverless cars and identify the most logical routes.

But what about AI in government? Business Minister Claire Perry announced £84 million of funding for AI and robotics research only recently. This was shortly after a government report explained that AI could add an additional £630bn to the UK economy, with the UK positioned to become one of the market leaders within the arena.

So, what is the potential of this technology when applied in government? Does the future look promising with it included? What role will it play? The possibilities are endless and will only ever evolve and develop with time. However, here are my initial thoughts on this.

The principle: cutting down costs

AI technologies can be particularly useful within the area in government that concerns administrative work. Why? Think about it - how much does the government currently spend on staff?

In December 2017, 5,350,000 million people worked in the UK public sector. Many of these roles are administrative and require hours spent on mechanical tasks.

According to Deloitte, low-investment AI (which speeds human tasks up by 20%) could free up 96.7 million human hours in government, while high investment AI (which speeds human tasks up by 200%) could save 1.2 billion human hours in government.

This could have potential savings of between $3.3 billion and $41.1 billion, or more likely, the workforce would be utilised in different jobs within the busy governmental administration.

The aftermath: Investing on weaker spots

Once administrative burden is reduced, the money can be spent on developing other AI-based government technologies to improve other weaknesses in the system, say for example healthcare.

We’re all aware of how overstretched the NHS is right now. AI is one way that it could run smoother. NHS leaders have said that patients’ illnesses could soon be diagnosed using AI.

But it’s not just diagnosis that can be improved with AI - artificial intelligence technologies can collect shared data which can create a “next-generation NHS” according to government officials.

Addressing challenges: the sensitivity issue

Government software and websites need to be smooth-running and glitch-free, considering they serve an entire population. For example, as the British public painstakingly learnt in 1992, when the software dispatching ambulances malfunctioned, there were serious consequences.

Consider the reduction of human errors as well as the vast amounts of work developers and testers would have on their plates if AI was implemented across the different array of governmental sectors.

So why not use AI to take over the testing tasks too? Software could guarantee more safety (avoiding any risks of human error), and once again, time and money would also be saved.

Integration: the fear of the unknown

This is not to say, however, that AI is without its limitations. One of the chief reasons AI is not being implemented at the rate it could be is the fear that surrounds the nascent technology: the fear of the unknown.

From a meta sense, cultures of fear are typically rooted in ignorance; the less you know, the more fearful you are. Indeed, according to a recent Sage survey, 46% of respondents in the UK confessed that they have ‘no idea what AI is all about’.

Once conflated with comments from industry titans such as Elon Musk who recently made the claim that artificial intelligence represents humanity’s greatest ‘existential threat’, we can forgive the populous for exhibiting a reluctance to push forward.

Looking beyond AI ‘doomsday’ scenarios however, we can find a viable concern that deserves our attention: automations’ impact on employment.

Let’s be clear - the implementation of AI-powered automation software will inevitably lead to job loss, both in the private and public sector.

An example of this is the rise of the chatbot, whose development (amongst other technologies) led analyst firm Gartner to predict that 85% of all customer interactions will be managed without a human representative by 2020.

So yes, many job types we have today will disappear, but jobs as such will not. The so called ‘fourth industrial revolution’ is all about helping humans do their jobs better; it performs the menial automation tasks we hate so we can get on with more strategic work.

Whilst we might need to retrain some of our workforce, AI is set to create a substantial amount of new job opportunities to us humans.

AI is here now - don’t be late

AI has huge potential in numerous areas, but it’s not all that easy to implement. What’s more, hurdles such as bureaucracy are compounded in the public sector.

The key - the conversation needs to shift from ‘man vs machine’ to ‘man plus machine’. AI is giving us the opportunity to automate and streamline our government services at a rate we have never experienced before.

By being flexible about it and allowing it to be incorporated into our processes we will be able to reduce costs and increase efficiency, our government will be able to properly concentrate its resources on frontline priorities.

Daniel Kroening is professor of computer science at the University of Oxford and CEO at Diffblue.