Redesign vs Replace: The Dilemma of Agentic AI
Back in March 2024 at the Dublin Talent Summit, I delivered a rally cry to +2,000 people and business leaders: “You have to start building an understanding of GenAI in your business and especially in HR - or risk losing HR’s seat at the table!”. It was the post-lunch-re-warm-up keynote and I intentionally injected a bit of drama; not just to address the mid-day slump, but specifically to warn and equip people leaders with what was to come … being front and centre to preparing organisations for a very new world of work in which our existing frameworks don’t fit.
Eighteen months on, the burning question of whether AI will replace jobs is still top of mind. Productivity figures have largely been unrealised. Because, the pace of change, AI adoption in this case, is only as fast as humans can absorb. This disparity is clearly seen through large-scale job cuts - but with quite mixed results:
- Shopify reports a 91% revenue increase with 30% fewer employees – currently limiting recruitment to “only hire what AI can’t do”.
- Microsoft has cut +16,000 roles since 2023 across several of their core competencies in product, engineering, gaming, sales and customer service, and yet posted +18% YOY growth.
- Cisco, Salesforce, Indeed announced 1,000–5,000 cuts, with the resulting financial gains offset by lower morale, candidate distrust, and stock volatility.
- IBM removed 8,000 mainly HR roles and still managed to double its NPS (net promoter score) - but is now rehiring due to poor employee response to automated HR services.
- Accenture has reduced its workforce by 11,000 in the past three months, warning staff more will leave if they “cannot be retrained for the age of AI”.
As the examples above highlight, we cannot deny that the AI horse has bolted - ChatGPT already exceeds one billion users, of which 700 million are active weekly – nor we deny that AI has and will lead to more job losses.
But are mass job cuts inevitable? To explore this more clearly, we must first remember every organisation’s AI journey is different, even from close competitors.
To understand the unique potential for your business, leaders must build AI know-how and literacy in themselves and across their business - AI tools give everyone the opportunity to do what they already do more effectively. And if you don’t understand the basics of what AI can and can’t do, you risk not being informed enough to make core decisions about the future of your business. Gen AI capabilities don’t sit tidily in the IT function (sort of like people don’t sit tidily under HR) and building AI literacy is a company-wide endeavour.
To dive into a highly relevant point on the human response to change, it’s important to remember we are biologically hard-wired as humans to resist change – the bigger the change, the more resistance naturally occurs and the more trust is needed to shift out of resistance. AI brings a greater emotional response to change than previous tech advances, and we’re right to pause to understand our responsibilities around ethical and risk-adverse approaches to AI. To build trust in AI, we need to give our employees a vision or at least a direction of travel on how AI is planned to integrate into the business, supported by an action plan to equip employees with the skills for safe and responsible use, and introduce policies providing clear direction on use of AI aligned to business needs and risk management.
To shift beyond ‘shadow’ AI use and create a bright light of AI literacy across the organisation, building trust is fundamental. Building trust is dependent on communication and genuine efforts to support employees through transitions and build psychological safety. Key questions employees want to know include:
- How will AI or Agents help in my role? How is my role changing?
- How will I be supported and equipped with the new skills I need to be successful as my role evolves?
- What’s our AI usage policy?
Redesigning vs replacing roles offers the chance to keep talent while unlocking new levels of productivity and innovation. It also supports our socio-economic model of apprenticeships. Juniors typically start with simpler work before progressing to higher-responsibility roles. As we said at the outset of this article, some jobs will go. We no longer plough fields by hand. But there’s a balance of thought for leaders to consider with the concept of Sunset and Sunrise roles – helping employees transition and upskill from those roles and tasks that are declining to those that are emerging.
In manufacturing, for example, freeing administrative staff from repetitive scheduling tasks can allow them to clean datasets and use them for predictive maintenance or production planning, or to investigate persistent quality issues, or work directly with suppliers on performance improvements. This is value-added work that rarely gets done when teams are buried in routine paperwork.
My closing thoughts for leaders in manufacturing and beyond: the challenge is not just adopting AI, but deciding how to reshape work so people and machines create more together. The organisations who pause to upskill and redesign, rather than replace, are the ones unlocking real and sustainable value.
Stephanie Prenderville is involved with Ibec Academy’s new HR Consulting service. To develop your organisation’s AI integration plan, please do not hesitate to get in touch.
Stephanie Prenderville
Founder
SPC Consulting
