Where is AI actually being used and creating value?

The last two years have been characterised by the meteoric rise of ‘AI hype’, but very little explanation of whether AI is delivering value and, if it is, where businesses are actually seeing the benefits. Michael Conway, the UK and Ireland lead for AI at IBM Consulting, tells Consultancy.uk what the technology is providing to British businesses, and how the firm is helping clients get the most from their early forays into it.

You’ve been in leading roles with IBM Consulting’s AI services wing since the end of 2020, when you became a partner. That’s some time before the public release of ChatGPT prompted the heightened levels of AI in public discourse. How has your role changed since?

Since stepping into a leadership role in Consulting's AI services in 2020, the AI landscape has evolved significantly, especially with the recent surge in AI discussions, partly driven by advancements like GPT. Over the years, my role has transformed from advocating for AI's potential to guiding organisations on how to harness it strategically. Today, AI has become a boardroom topic, sparking conversations around innovation, efficiency, and long-term competitive advantage.

As a technology optimist, I see this shift as incredibly exciting. AI has opened up new possibilities in decision-making and problem-solving, enabling businesses to move from proof of concept to real, measurable outcomes. With our tenure in AI, we're well-positioned to expedite these outcomes. The years of expertise IBM has built serve as a strong foundation, allowing us to cut through the noise and focus on creating tangible value for clients.

While there is certainly more pressure due to the growing expectations and the rapid pace of AI advancements, I also find the current moment exhilarating. The demand for AI-driven innovation has grown, and with it, the responsibility is to deliver strategies and solutions that are ethical, sustainable, and transformative. Ultimately, the excitement outweighs the pressure, as we now have the tools and experience to deliver meaningful impact on a global scale.

IBM experts have insight into the entire AI journey and all of the challenges organisations face along the way – from governance, security and skills to sustainability and AI strategy. What are the most pressing challenges you are helping clients with on their AI journey?

The most pressing challenges we are helping clients address on their AI journey centre around scaling, governance, and the ever-changing regulatory environment. Scaling AI from pilot projects to full-scale enterprise solutions is a major hurdle for many organisations, often requiring significant adjustments in infrastructure, talent, and operational processes. Governance is another critical area, where organisations must ensure that AI is used responsibly, in compliance with emerging regulations, and in ways that promote trust and transparency.

Security is equally important, as integrating AI into business processes introduces new vulnerabilities that need to be addressed to protect data and systems. Ultimately, building trust – both within the organisation and with external stakeholders – is essential, as transparency and ethical AI use are now vital to sustaining AI’s long-term value.

There are many practical possibilities which have been mentioned around AI, but the one that seems to most commonly talked about is Generative AI’s impact on content creation. One example of that came when IBM partnered with Wimbledon, using a GenAI system to produce ‘editorial’ content for the tennis competition. But the project was not without its teething problems – and it did not take social media long to pick up a number of errors which it published. Do you think the public nature of many of these explorative projects is helpful, or is it damaging the perception of AI as a useful tool?

When creating new AI applications, there are always learnings; not everything will go perfectly. The issue on the first day of the Wimbledon was fixed very quickly and worked fine for the entirety of the tournament. What’s important is that the right governance is in place (watsonx.gov) and there’s a human in the loop.

It has been said that there is potential for AI to make the UK’s public sector more efficient and productive. However, recent research from the Australian government suggests that – in its current form – GenAI may actually make more work for employees rather than less. Should governments be looking before they leap, when it comes to adopting the technology in pursuit of savings? Or do they just need to find the right use-cases?

We are already seeing growing interest and appetite from the UK’s public sector to embrace GenAI for productivity and business transformational needs. Our work with the central government has already highlighted the growing benefits of AI adoption outside of cost saving such as improving citizen services, enhancing decision-making, and increasing operational efficiency.

Following on from that, and highlighting other ways AI is actually being used, IBM is already working with the NHS to use AI to reduce waiting lists. What exactly is being done, and what tangible results has that yielded so far?

IBM has been collaborating with University Hospitals Coventry and Warwickshire (UHCW) to address missed outpatient appointments. By analysing data on appointment no-shows, we identify patterns and implement interventions to reduce missed appointments and improve patient care with the use of IBM’s iX garage methodology, data mining and AI. The impact saw appointment “no shows” decreased from 10% to 4% – a 60% improvement. An additional 700 appointments became available per week, equal to more than 36,000 throughout the year.

The energy use of AI remains a bone of contention, though. How would those savings compare to the embodied carbon used by the systems which do the machine learning?

We aim to assist our clients in achieving their sustainability goals by providing AI-powered solutions and strategies that can help reduce energy consumption, improve resource optimisation, and facilitate faster decision-making related to our client's sustainability initiatives. The key imperative is to use the right approach to solve the business problem. This leads to efficient use of technology to solve a problem, rather than looking for a nail because you’re holding a hammer.

Following on from that, AI is talked about as a ‘double-edged sword’ when it comes to fighting climate change. On the one hand, AI modelling has been touted as a way in which scientists and governments could use to make important decisions about emissions. But some experts have also warned that the huge amounts of energy that the technology relies upon – at a point when the world is still struggling to meaningfully transition away from fossil fuels – means it could also expedite global warming. How can AI providers account for this, enhancing the benefits of machine learning, while lessening the potential harm?

To minimise AI's environmental impact, it is important to make smart choices at every phase – training, tuning, and inferencing. Opt for a mix of foundation models, which are appropriate for the task at hand and can be tuned for specific purposes. Use a hybrid cloud approach, balancing energy efficiency by processing tasks in data centres near the demand or on-premises for security and regulatory reasons.

It was recently suggested by Jesse Dodge, senior research scientist at the Allen Institute for AI, on The Bunker podcast, that AI developers are not being transparent enough with organisations adopting the technology’s emissions. But firms need to understand those impacts when it comes to their net zero journeys? How could firms and governments prepare for this?

This is a nascent market and it is critical that organisations understand the carbon impact of the models they are consuming. Over time, we expect businesses and governments working towards carbon neutrality to understand the impact of AI on their net-zero journey. Using open-source AI models, organisations can customise and adapt these solutions to their unique challenges, which tends to be a more efficient approach when it comes to carbon. Transparency is the name of the game though, knowing the full value chain impact of carbon impact is going to be critical for organisations to understand.

For some time now, market commentators have been suggesting that the huge investments flooding into AI development are a ‘bubble’, and that it is ripe for bursting. With chipmaker NVIDIA having $279 billion shaved from its stock value in a single day, and backers of the technology like Microsoft and Meta also said to be haemorrhaging money on AI projects, is the market entering that long-prophesised period of correction, or is the situation more complicated than that?

While market fluctuations and high-profile corrections can create the impression of a bubble, I remain confident that enterprises will see significant long-term benefits from AI. These short-term market movements don't reflect the real value that AI is delivering across industries. As the technology matures, we're seeing businesses unlock new efficiencies and innovation, which will continue to drive sustainable growth. The situation is more complex than a simple "boom and bust" cycle – AI is evolving.

Do you expect that could lead to another ‘AI Winter’ that could hold back investment on AI in the future? And would IBM Consulting’s plans shift in that case, when it comes to AI integration services?

Not at all. At IBM, we have been at the forefront of change, investing heavily in research and development of advanced AI technologies, such as foundation models, and collaborating with leading institutions like NASA to push the boundaries of what's possible. We have also been internally invested in AI with platforms such as IBM Consulting Advantage, providing our consultants with secure, easy and customised access to our AI assets and models and role and task-based generative AI assistants and proprietary methods.

We strongly believe that the business world will reorient around this technology, so we are transforming ourselves into “client zero” and working with our clients to maximise their opportunities.

Finally, a lot of the discourse around AI has dealt in absolutes. It is either here to transform the world, or it is an utter waste of time and money. If the technology is to have a sustainable future, without a funding system that is prone to booms and busts, does the dialogue around AI need to move beyond the ‘hype cycle’, and concede that the technology is still a long way away from its apparently revolutionary potential, but could still be useful in the long-term?

The hype cycle around AI is very much real, but it's important to move beyond it and recognise the tangible business value AI is already delivering today. While the technology may still be developing toward its full potential, we are already seeing real-world applications that drive efficiency, innovation, and competitive advantage. Our work with Natwest CORA and UHCW NHS Trust are testament to our client’s success.

At IBM, we've been at the forefront of AI adoption much longer than others, and our focus has been on guiding our clients through this journey, helping them harness AI's current capabilities while preparing for future advancements. The key is to approach AI as a long-term investment, not a short-term one. By focusing on sustainable, practical use cases, we can help businesses realise immediate benefits while staying on track for the larger, transformative potential of AI in the years to come.