Technology tends to outpace the ability of the average user to understand how it works, and AI is no exception: today's models are popularly dubbed 'black boxes' for the opacity of the algorithms used. This gap between function and user understanding gives rise to plenty of uncertainty around how to use AI and AI-derived technology, when and where to use it, and even whether it is safe to adopt or implement.

We picked up a few pointers to help answer these questions at a recent thought leadership event organised for HR leaders by EY and SAP with the support of People Matters. Jason Seng, partner in EY's People Consulting practice, provided some definitions and answers to help demystify AI and AI's role in HR.

What is AI, really?

Seng suggested a basic definition of artificial intelligence for those without a technical background: a computer-based intelligence developed by humans, characterised by speed, accuracy, consistency, and reliability. To that, he added a simple breakdown of three frequently seen terms.

Deep learning: the process by which an AI model learns to recognise patterns by evaluating a large volume of content. Facial recognition systems are an example of models that use deep learning.

Generative AI: an AI model that is able to generate new or new-appearing content after evaluating a large amount of content. ChatGPT is the most well-known example of such models.

Conversational AI: an AI model that is able to replicate the experience of conversing with a human, like Apple’s Siri or Amazon’s Alexa. Chatbots are one of the most common examples of such models.

So what can AI do? Quite simply, he said, it can streamline certain transactional activities from people's workday, potentially reducing some aspects of their roles. This has a knock-on effect across the workplace, where organisations have to respond by either transforming jobs to align with new technologies or adjusting staffing levels - hence the common fear that AI will take away jobs - or ideally, changing the nature of jobs.

Can AI replace people? The answer is: not completely

It's the nature of jobs that need to change, and this is why, Seng says.

Human intelligence - which he describes as practical intelligence - is characterised by its roots in life experience and social learning, he said. It has aspects that artificial intelligence does not replicate, including creativity, emotional intelligence, and moral and ethical decision-making.

What that means, he said, is that AI cannot replace people at work, and it certainly cannot replace HR. In areas such as strategy and development, complex negotiations, or personalised coaching, human judgement continues to be the only way to achieve high quality outcomes.

But AI can take the role of a personal HR advisor: a robot, so to speak, sitting beside HR practitioners and helping them to shape performance goals or create better accommodations for employee experience.

In particular, Seng said that AI empowers HR and employees, allowing them to surface and address needs autonomously without needing to wait for leadership intervention.

"It's not about applying AI to your job but to your personal career, shifting your capabilities, improving your eligibility for promotion," he pointed out. "The better humans can use AI, the more effective humans can be."

A framework for AI adoption

"I believe that AI presents the greatest opportunity today -that allows HR leaders to step forward and effect organisation-wide impact," said Seng. "HR has always been aspirationally trying to move to become a strategic partner with a  seat at the table, but I think AI can help HR leaders take another step up and become transformative leaders."

He suggested a holistic framework that brings together various key elements: resilient culture, change management capabilities, organisational structure, efficiency-focused processes, and robust governance. Importantly, organisations that want to adopt AI have to bear in mind that every business and in fact every function has its own specific needs, and the use case has to be differentiated by needs to achieve the best ROI.

"Think about a mindset shift," he advised. " AI is not primarily a cost cutting tool. It's about making you more productive and more effective as an individual performer. It's about getting people to focus on value-added activities because they don't have to do the grunt work."