New Maps for a New World of Work
A conversation with Jeff Schwartz, Episode 7 of Workestration
Einstein said you can't use an old map to explore a new world. Jeff Schwartz opened our conversation with that line, and it framed everything that followed.
Jeff has spent fifteen years studying how work changes when technology changes. He wrote Work Disrupted (Wiley, 2021) and co-authored Workforce Ecosystems (MIT Press, 2023) with colleagues at MIT Sloan and Deloitte. He teaches at Columbia Business School, leads Insights and Impact at Gloat, and was named a Thinkers360 Top 50 thought leader on the future of work. He came to the podcast with a clear thesis: the institutional models we inherited from the 1980s and 90s do not stretch to fit what AI is actually doing inside organizations right now.
A few ideas from the conversation are still lingering in our minds.
The false binary
Industry has been debating whether the future of work is AI-first or human-first. Jeff thinks the framing is wrong. The Burning Glass Institute published a report called "Beyond The Binary: How Automation and Augmentation Are Combining to Reshape Work" which found that automation and augmentation are happening in the same jobs at the same time. Parts of a role get handed to machines while other parts get amplified by them. The radiologist who Geoffrey Hinton predicted would be obsolete in 2016 is, according to recent reporting on the Mayo Clinic, working in larger numbers than ever, alongside AI, in subspecialties like interventional radiology that did not exist a decade ago. Hinton updated his view in 2025, conceding he had read the technology correctly but misread the work itself. Eric Topol writes about this in his books Deep Medicine (Basic Books, 2019) and Super Agers (Simon & Schuster, 2025).
Adoption is mostly done. Redesign has barely started
Deloitte's 2026 State of Generative AI in the Enterprise report found that 84% of companies have not begun to redesign work around AI. The first phase, getting employees comfortable with tools, has gone reasonably well. The second phase, asking what the work even is when machines can do parts of it, sits on every leadership team's desk. Jeff put it bluntly: "If you were asking in 2026 what transformation initiatives you should be undertaking for 2030, and it is not highly disruptive, you're probably missing the mark." Four years ago, ChatGPT did not exist. OpenAI has shipped 21 models since then. Thirteen are already retired. The clock is moving.
The trap of incremental thinking
It is tempting to apply AI to whatever you already do and call it transformation, but Stela kept pressing on the harder question: whether the work needs to happen at all. Michael Hammer and James Champy raised a version of this in their 1990 Harvard Business Review essay "Reengineering Work: Don't Automate, Obliterate." Most organizations took the easier path then, optimizing what they already had. They are at risk of repeating the move now.
HR as work architect
If the workforce includes employees, contractors, freelancers, partners, AI agents, and algorithms, then the function that organizes them needs a different remit. Jeff suggested HR teams stop asking how to improve their existing processes with AI and start asking what HR even is when "workforce" includes machines. He pointed to Tracy Franklin at Moderna, whose remit reaches across HR and parts of technology. Not a template for every company. Worth watching for what it signals about treating workforce composition, work design, and reskilling as a single discipline.
Managers as designers
Most management training still assumes the job is to assign work, monitor compliance, and evaluate performance. In a world where some of that work is done by an algorithm sitting between the manager and the worker (think Uber's dispatch system, where the algorithm manages drivers in real time), the human manager's value moves elsewhere. Jeff described the shift as supervisor to designer, evaluator to coach, compliance officer to behavioral economist who builds environments where people thrive. We will be writing more about this.
Transition nets, not just safety nets
Jeff closed on what keeps him up at night, and the answer was policy lag rather than the technology itself. If people are going to make four or five career transitions in their lifetimes, the public infrastructure has to support that. Income support during retraining. Portable healthcare. Pension contributions that follow the worker. Europe has been working on this for years through systems like Denmark's flexicurity model. The US has catching up to do.
What we took away
Two thousand days from now is 2030. Most of the leaders we talk to are still treating AI as a tools rollout. That phase is largely over. The work that creates real value sits on the other side of a redesign question that 84% of organizations have not yet asked.
The Sherpa metaphor Jeff used at the end of the conversation is the right one. None of us has been up this mountain before. The best we can do is share what we have seen with the people climbing alongside us.
That is most of why we started the podcast.
Listen to the full conversation with Jeff Schwartz on Episode 6 of Workestration. Subscribe wherever you get your podcasts.