Skip to main content

Q&A: How can Artificial Intelligence change transformation?

(Image credit: Image Credit: John Williams RUS / Shutterstock)

In the same month it was announced that computer pioneer, codebreaker and the father of Artificial Intelligence (AI), Alan Turing, will feature on the new design of the Bank of England's £50 note, a report from IDC suggests that spending across Europe will hit $5.2 billion in 2019 up from 2018 by 49 per cent and expected to rise to $13.5bn by 2022.

The race to build machines capable of matching and beating human vision, language understanding, intelligence, and movement is on. Change management and business transformation is one area of business operations where AI could make an exciting impact. In this article, three experts share their thoughts to help to better understand what AI is, how much of what we know is real, what it can and should do, and, most importantly, how it might affect business transformation programs around the world.

BN: We often see the terms AI mixed with automation, deep learning and machine learning, what do they all mean?

Matt Armstrong-Barnes, Chief Technologist for Artificial Intelligence at Hewlett Packard Enterprise: Unfortunately there is quite a lot of confusion around AI at the minute, mainly because it is such a big discipline. If we think about AI as two main streams there is strong AI and narrow AI.

Strong AI is the stuff of science fiction, with machines operating at the same level as humans or, to put it differently, machines that have the ability to apply intelligence to lots of problem domains and a broad set of problems; we are a long way away from achieving this.

Narrow is where AI is today and it is really about using machines to focus on a narrow or very specific problem space, but a problem space with lots of data and rules, so complex they cannot be written down or if they can they are virtually impossible or very costly to update / maintain. This means that AI today is focused on a specific set of use cases -- and it’s here and achievable now.

Narrow AI is really achieved using a technique called Machine Learning which is a machine capable of changing its behaviour based on historical information or identifying patterns that it has not been explicitly programmed to spot. For example, if you think of spam filters, they can be used to block email when you told them that an email was spam, the machine would 'learn' what to look for and would then block the next time without involving the user. The most popular way of achieving Machine Learning is Deep Learning. This technique, inspired by biological brains, uses a mathematical model to represent artificial neurons and create an artificial neural network to make a machine appear intelligent.

To think about it in a slightly different way, AI covers anything which enables computers to behave more like human brains. ML is the subset of AI that deals with the extraction of patterns from data sets. This means that the machine can find rules for optimal behaviour but also can adapt to changes in the world; it can learn. What this means is machines can find rules in subject matter that the manual creation of would prove either extremely costly or practically impossible.

BN: The tech world often sees technologies come and go -- hype and hyperbole are common. Is there is risk that AI is the next fad or more vapourware?  

Armstrong-Barnes: AI -- as a term -- was first coined in the 1950s, and that in turn came around from work done in the 1940s looking at missile trajectories. The mathematics behind AI has been around for a very long time, you could argue that AI has its roots in formal reasoning, which goes back hundreds of years.

It has also had the support of large businesses for quite some time. Back in 2012, a computer scientist called Alex Krizhevsky significantly won the annual ImageNet Large Scale Visual Recognition Challenge [or ImageNet competition] using AI and a mathematical model powered by GPUs, rather than CPUs.  The model he created recognised more images significantly quicker than his fellow competitors. Realising the potential of combining AI and GPUs, chip manufacturer Nvidia was soon involved -- followed by many other large vendors such as HPE, Amazon and Google.

Ryan Tabberner CEO of business transformation consultancy, Servita: When you consider that Alan Turing dreamed of machines to rival humans nearly seventy years ago, AI is an established technology, but perhaps new to business transformation.

We’re constantly exploring new ways to help our clients transform their businesses and AI is an area that we see more and more companies investigating and adopting. It will be here to stay as long as clients focus on their business objectives, and build a clear transformation program first and foremost, and don’t expect AI to win it all. It needs to be a tool in the tool box not a panacea for everything.

Chris Leary, UK Regional Director at Servita: From my perspective it’s clear that AI is here to stay and can help make organisations more competitive and efficient i.e., the ethos of transformation. More and more companies will seek to exploit the value it can offer their businesses. As they understand what it can do for them, it will become a strategic growth factor.

BN: What can AI really deliver; what can it and should it be doing?

Armstrong-Barnes: At HPE, we work with customers wherever they are on the AI journey. From companies who are just thinking of starting to mature companies who are already working extensively with AI. Current analysis from IDC shows the big spenders are in retail, banking and financial services, healthcare and manufacturing. As a big manufacturer ourselves we use AI in our own processes to improve quality and schedule downtime more effectively.

BN: Plans are also taking shape in the UK to launch a new public sector procurement framework outside of G-Cloud specifically for the areas of machine learning, artificial intelligence, analytics and robotic process automation. Does this mean the public sector is a big market too?

Tabberner: The UK Government is throwing its full weight behind AI. In May this year it unveiled its AI Council which includes Google, Microsoft and Amazon as members. The Council will look at ways to adopt AI across the British economy. The Government has previously established an Office for Artificial Intelligence too, which manages government policy for the sector.

Armstrong-Barnes: We are working with a UK Police Force at the minute which is wrestling with the challenge of using their resources in the most effective way. They didn’t come to us asking for an AI, but due to the complexity of the information they were processing, AI was the only answer. So, we built an AI to predict if the outcome of an arrest would lead to a conviction. AI is a tool that should be in every organisations toolkit and, in this instance, it is used to help investigating officers sift through large and complex datasets and make more informed decisions.

The NHS, as part of the Long Term Plan, has highlighted that AI is going to be a significant way that the NHS can be more effective in patient care. Again, an example of increasing demand and limited resources. This has further been enhanced with the announcement that a new AI Lab will be formed and £250m allocated to it to make the NHS a world leader in AI by 2024. AI is seen as a critical technology to speed up cancer diagnosis and revolutionise outpatients.

Leary: Beyond the headlines in the NHS, there are other areas were machine learning is being applied to good effect. AI can streamline case management, ensuring customer support agents receive real time advice as to how a given query should be handled. Chatbots provide intelligent answers to online queries without the end user having to wait for an agent to become available. Large datasets, containing many millions of annual transactions and hitherto unfathomable, become rich sources of insight through the application of machine learning tools and data visualisation products.

BN: Lots of the examples appear project based. How can AI help to transform whole organisations?

Armstrong-Barnes: AI needs to be addressing a business problem to be successful, I see cases where organisations are implementing AI for technology's sake and these are interesting science projects but do not support the business value that AI can bring. We have to remember that AI, or rather Deep Learning, is a mathematical model, so it needs access to the right data in order to be successful. It is best incorporated as a cornerstone into wider program of change or transformation. To make change in an organisation work, people need to have information and they need insight, AI can help to do this. AI is about three simple steps: explore, experiment, evolve.

BN: Should companies be upskilling workforces or should they enlist the help of external partners?

Armstrong-Barnes: There are two ways of approaching an AI implementation, either skill up and go on the journey yourself of engage with a partner who has done it before. The quickest way of getting started is to partner. There are lots of great organisations to help identify use cases, the technology needed and potential data challenges. Ultimately, clients need to find a partner that they can go on a journey with.

From a HPE perspective, we are definitely about working in a partnership and engaging with our customers and ecosystem of partners to help bring the right value to the right problem. We want to leave our customers with the skills they need to be successful in going on the AI journey, because it is a journey not a destination.

Leary: From Servita’s perspective, AI is another 'technology' so the challenge for companies is around getting from where they are today to harnessing the value of AI. They’ll need lots of help to understand the benefits that AI can bring; the organisational changes needed, adoption plans, management changes and the like.

Ryan Tabberner, CEO Servita
Christian Leary, UK Regional Director,
Servita
Matt Armstrong-Barnes, Chief Technologist for Artificial Intelligence,
Hewlett Packard Enterprise