Mike Sullivan, Vinay Gidwaney - Headcount Is the Wrong Number - What HR Should Measure Once AI Joins the Team

David, Mike and Vinay discuss how HR can measure work as AI becomes part of the workforce. They argue that headcount alone misses the shift: organizations need to understand which tasks people and AI handle, how AI changes people’s capacity, and how to use that capacity to create more value without losing the human contribution. 

Key topics and takeaways 

• Measure work, not just headcount. The guests describe “workforce intelligence” as a way to manage and measure a blended workforce of people and AI coworkers. 

• More capacity doesn’t automatically mean less busyness. Vinay says employees may get time back from AI but feel pressure to do more. Organizations need to help direct that “capacity dividend” toward valuable work. 

• AI output still needs human oversight. The host points to extra quality checks, hallucinations and the volume of AI-generated work as reasons that adoption may not immediately save time. 

• Efficiency is only part of the opportunity. Mike says it is easier to frame AI as a way to reduce costs than to work out how people can use it to deliver new, distinctly human value to clients. 

• Treat AI as a workforce cost. Vinay says OneDigital assigns each AI coworker a salary to understand its operating cost. The guests discuss matching model capability and cost to the task, rather than using the most advanced model for everything. 

• Rethink how work is organized. Mike says organizations should look at the tasks that make up jobs, assign work to the right “unit,” and help employees adapt. He estimates the coming challenge will be mostly organizational change and adoption, rather than technology alone. 

• Human accountability remains important. Vinay argues that AI can make information more accessible, but human judgment, experience and accountability are what distinguish an organization. 

• Make AI learning part of education and work. Vinay encourages using AI to expand thinking, not just to get answers with less effort. He says people who avoid learning to use it may miss career opportunities. 

• People’s AI journeys differ. Mike describes AI as a source of new ideas and connections, while Vinay stresses that organizations need to understand what adoption means for individuals—not assume one approach suits everyone. 

Names and resources mentioned 

• OneDigital — Mike and Vinay’s organization. 

• Work Carousel — episode sponsor; https://workcarousel.com/join/ 

• AI tools and models: Claude, ChatGPT, Copilot, GPT-OSS, and open-weight, non-frontier and frontier models. 

• Hardware: NVIDIA DGX Spark. 

• Other platforms and organizations mentioned: Microsoft, Fiverr, Deloitte, Accenture, Instagram, Penn State. 

• Terms discussed: workforce intelligence, AI coworkers, centers of intelligence, tokens, and retrieval-augmented generation (RAG).