Ling-Yi Tsai, our HRTech expert, brings decades of experience assisting organizations in driving change through technology. She specializes in HR analytics tools and the integration of technology across recruitment, onboarding, and talent management processes. Today, she unpacks the nuanced reality of AI’s integration into the workplace, moving beyond the headlines of job replacement to explore a future defined by human-machine partnership. The conversation touches on why uniquely human skills are becoming more valuable, how a background in the humanities offers a competitive edge in a tech-dominated world, and how leaders can navigate the tension between AI as a supportive tool and a disruptive force.
Powerful AI models can now write a company’s internal code, sparking fears of replacement. Why do you believe the number of fully autonomous AI jobs remains minimal? Could you walk us through a specific scenario where human partnership is still essential for success with these advanced tools?
It’s a fear I see in boardrooms constantly, but the reality is that the number of jobs AI can perform fully, without any human guidance, is vanishingly small. The technology works best as an incredibly powerful partner, not a substitute. Think of it this way: a tool like Claude Opus 4.6 can write nearly all of a company’s internal code, which is an astonishing feat. But who defines the problem it needs to solve? Who understands the unique culture of the company and the nuanced needs of the end-user? A human programmer has to provide the initial vision, the strategic direction, and then critically evaluate the AI’s output. They are the ones who must integrate that code into a complex, existing ecosystem, troubleshoot unforeseen conflicts, and, most importantly, provide the creative spark that leads to true innovation. The AI is the ultimate assistant, but the human is still the architect.
At Anthropic, you prioritize humanistic skills like emotional intelligence and compassion over purely technical credentials. What does that look like in practice during the hiring process? Can you share a specific interview technique or question you use to assess these qualities in a candidate?
This is a critical shift we’re seeing across the most forward-thinking tech companies. Moving beyond a resume packed with technical acronyms is essential. In practice, this means structuring interviews around behavioral and situational questions that reveal a person’s character. For instance, instead of asking someone to solve a coding problem on a whiteboard, we might ask them to describe a time a collaborative project failed. We aren’t looking for them to blame others; we want to see if they can demonstrate self-awareness, curiosity about what went wrong with the team dynamics, and compassion for their colleagues’ perspectives. We listen for how they communicate the story—do they show emotional intelligence in their analysis? That tells us far more about their potential to thrive in a collaborative, AI-driven environment than just their ability to write code.
You’ve argued that studying the humanities is becoming more important as AI excels at STEM tasks. How has your non-technical background in literature specifically benefited you in the AI industry? Please provide a concrete example of how it has shaped a business or product decision at Anthropic.
My background in literature has been invaluable, precisely because these AI models are already so proficient at STEM tasks. Literature is the study of the human condition—what motivates us, our histories, our flaws, and our triumphs. When you’re building technology that will interact with people on a massive scale, that understanding is not a soft skill; it’s a core competency. For example, when we were developing the communication protocols for a new AI assistant, my experience with narrative and rhetoric was crucial. I argued that the AI’s responses needed to be more than just accurate; they had to be built with an understanding of context and empathy. This led to a product decision to invest heavily in training the model on different communication styles, ensuring it could adapt to a user’s emotional state, a concept drawn directly from understanding character and dialogue in literature.
While you emphasize AI as a supportive tool, your brother and CEO, Dario Amodei, has expressed concern that AI’s labor impact could be faster and broader than past technological shocks. How do you reconcile these two perspectives, and what steps is Anthropic taking to manage this potential disruption?
These two perspectives aren’t contradictory; they are two sides of the same coin and represent a responsible outlook. My focus is on the nature of the work in an AI-integrated future, which will absolutely be a partnership. Jobs will become more meaningful and challenging as humans offload rote tasks to their AI counterparts. Dario’s concern, which I share, is about the velocity and scale of this transition. Past technological shocks, like the industrial revolution, affected a fraction of human abilities, leaving people room to adapt and find new tasks. This shock could be different. The way we manage this is by leading the conversation, being transparent about the technology’s capabilities, and fundamentally rethinking our hiring philosophy. By prioritizing adaptable, critically-thinking, and emotionally intelligent people, we are building a workforce that is resilient enough to navigate this disruption, turning a potential crisis into an opportunity for more engaging work.
What is your forecast for the future of work?
My forecast is one of profound realignment rather than replacement. The future of work won’t be about humans versus machines, but about humans amplified by machines. The most valuable professionals will not be the best coders, but the best collaborators, communicators, and critical thinkers who know how to leverage AI to solve complex problems. We will see a necessary and long-overdue revaluation of skills rooted in the humanities—empathy, ethical reasoning, and historical understanding—as they become the essential differentiators in a world where technical execution can be automated. The transition will be challenging, and it will intensify competition, but the result will be jobs that are more engaging, more creative, and ultimately, more human.