I’m thrilled to sit down with Ling-Yi Tsai, a seasoned HRTech expert with decades of experience guiding organizations through transformative change via technology. With a sharp focus on HR analytics and the seamless integration of tech in recruitment, onboarding, and talent management, Ling-Yi has a unique perspective on how AI is reshaping workplace dynamics. In this interview, we dive into the evolving relationship between humans and technology, the pitfalls of outdated terminology, the risks of over-trusting AI, and the innovative strategies needed to foster responsible collaboration in an AI-driven world.
How did the concept of the ‘end user’ come about, and why do you think it’s becoming outdated in the context of AI?
The term ‘end user’ originated in the 1960s from systems engineering, where it described non-technical folks who were simply operating a finished tool at the end of a linear design process. Think of it as the last stop in a waterfall model—value flowed one way, and these users were passive recipients. Back then, it made sense because technology was rigid, and roles were clearly defined. But with AI, that framing falls apart. AI isn’t just a tool you use; it’s a system you interact with, challenge, and shape. It demands active engagement, not passivity. Calling someone an ‘end user’ today ignores the complexity of decision-making and responsibility that AI introduces, especially in dynamic, collaborative environments.
What are some of the dangers you see in clinging to ‘end user’ thinking when working with AI systems?
The biggest danger is that it fosters a mindset of blind trust in technology. When we label someone an ‘end user,’ it implies they’re just receiving outputs, not questioning them. With AI, that’s risky because of how prone we are to automation bias—over-relying on systems even when they’re wrong. This can lead to serious errors, like lawyers citing fake case law generated by AI or employees publishing unverified content. It’s not just about mistakes; it’s about shirking accountability. If we keep this outdated mindset, we fail to recognize that we’re not just using AI—we’re responsible for its outcomes.
Can you explain what automation bias is and why it’s such a critical issue in the age of AI?
Automation bias is this human tendency to trust automated systems too much, even when evidence shows they’re off base. It’s a big deal with AI because these systems often present outputs with such confidence that we assume they’re correct. Unlike older tools, AI can produce errors or hallucinations that aren’t immediately obvious, and if we’re not vigilant, we accept them as fact. This isn’t just a minor glitch; it can lead to real-world harm, like flawed business decisions or legal missteps. It’s critical because AI is woven into high-stakes processes now, and over-trust can amplify the consequences of its flaws.
How should AI influence the way we view our roles when interacting with technology?
AI forces us to rethink ourselves not as mere users but as collaborators. It’s not about pressing buttons and getting results; it’s about partnering with a system that learns, adapts, and sometimes errs. This shift means we need to actively shape and question AI outputs, not just accept them. It’s a mindset change—seeing ourselves as co-creators of outcomes. Accountability becomes central here. We have to know where our responsibility begins and ends, ensuring we’re not just deferring to the tech but owning the decisions it informs.
You’ve pointed out that terms like ‘agent boss’ still focus too much on the tool itself. How can we better define our roles in relation to AI?
Terms like ‘agent boss’ keep us tethered to the technology rather than emphasizing our human responsibility. A better approach is to define roles based on what we do with AI, not how we manage it. For instance, focusing on decision-making authority or stewardship of data shifts the conversation to our accountability. I’d advocate for language that highlights purpose—like ‘decision steward’ or ‘outcome driver’—to underscore that we’re not just overseeing a tool but guiding its impact. This reframing centers us on our judgment and influence, not the tech.
Can you walk us through the idea of replacing ‘end user’ language with role-based precision and why it matters?
Role-based precision means ditching the generic ‘end user’ label for specific identifiers that reflect how people actually interact with AI. For example, you might have an ‘agent trainer’ who teaches the system, a ‘data steward’ who ensures input quality, or an ‘output validator’ who checks results. These distinctions clarify responsibilities and make accountability tangible. It matters because vague terms lead to vague expectations. When you name roles precisely, training becomes more relevant, requirements are clearer, and rollouts are smoother. It’s a foundation for responsible AI adoption.
What are interaction maps, and how do they help in navigating an AI-driven workplace?
Interaction maps are visual tools that show how different roles engage with AI and each other in a system. Unlike data maps, which just track information flow, interaction maps highlight human actions—who interprets outputs, who approves decisions, who escalates issues. They’re crucial in an AI-driven workplace because they make visible the complex web of collaboration and accountability. By mapping these dynamics, organizations can spot gaps in oversight or training needs. It’s about empowering people to manage AI actively, not just consume its outputs passively.
What’s your forecast for how AI will continue to reshape our understanding of roles and responsibilities in the workplace?
I see AI pushing us further away from static roles toward fluid, responsibility-driven identities in the workplace. As systems become more modular and agent-based, we’ll need even greater clarity on who holds authority and accountability at each touchpoint. I predict a growing emphasis on continuous learning and adaptability, with roles evolving based on how we interact with AI, not just what we do. Organizations that embrace this—by naming roles precisely and fostering active collaboration—will thrive. Those stuck in old ‘user’ mindsets risk falling behind, both in innovation and in managing the ethical challenges AI brings.