McKinsey’s updated Rewired playbook argues that AI delivers business value only when organizations rewire people, processes and leadership—not merely deploy new tools. Dan Swan, senior partner and global leader of McKinsey’s Tech & AI group, spoke with WIRED Global Editorial Director Katie Drummond about what separates companies that create measurable value from those that only experiment.
Focus on two or three end‑to‑end domains
Swan emphasizes focus: successful organizations choose two or three domains—end‑to‑end business areas or processes—where AI can drive systemic change rather than scattering small pilots across the company. Historically many pilots were isolated “use cases” that never scaled or appeared on the P&L, causing frustration. Reimagining a domain bundles multiple use cases and can be large enough to move the financials.
It’s a business transformation, not just a tech buy
The crucial shift is asking “what business problem do we need to solve?” rather than “where can we apply AI?” That reframing connects what the technology can do with what people must do and what organizational capabilities are required. Treating AI as a broad business transformation produces different outcomes than treating it as a new tool you point at various tasks.
Leadership, apprenticeship and practical questions
Swan advises leaders to ask concrete questions—did you use the LLM to check this? which tool did you use?—and to have senior leaders engage directly with more junior colleagues to learn how tools are used. Apprenticeship, judgment and communication remain essential as organizations adopt new AI capabilities.
Lessons from Toyota and DBS
- Toyota: A VP asked, “If Amazon can track a $10 book, why can’t we track a $40,000 car?” Toyota built a digital twin of its entire supply chain—from parts procurement through manufacturing to delivery for about 2.5 million cars. The digital twin could process a billion decisions in 15 minutes. What had required 100 people and 75 spreadsheets was consolidated into one solution, creating roughly $800 million in value via increased throughput and better timing. Crucially, sales, manufacturing and planning experts were involved from day one, which built trust and produced internal change agents.
- DBS: The bank treated AI as a fundamental transformation from the outset. By starting with two or three priority domains and then expanding to four to six as capabilities compounded, DBS moved faster and widened its competitive advantage. DBS’s prior experience with tech transformation and a people‑first approach meant the organization was receptive and ready to scale.
Agentic AI: new processes and new human roles
Pre‑agentic AI often sped up existing processes. Agentic AI enables autonomous agents to interact and form closed‑loop processes, changing how work is organized. McKinsey describes placing humans “above the loop” rather than “in the loop”: humans provide context, orchestrate and check the system rather than manually verifying every step.
Swan outlines a four‑rung agentic ladder: 1) asking an LLM a question, 2) embedding LLM features in existing software (e.g., meeting summaries), 3) fundamentally transforming a process, and 4) building a new business. Many companies are stuck on rung two because it is democratized, immediately useful, and requires few hard organizational decisions.
Managing AI slop
Successful organizations are intentional: they give agents context, measure inputs and outputs clearly, monitor hallucinations, and continuously improve. With this discipline, agents can learn over time and the amount of “slop” in the system can be reduced substantially.
Building capabilities and the role of HR
Swan highlights three HR priorities: setting the right narrative and mindset for transformation; building new skills—because required capabilities change rapidly; and redesigning how work gets done. Training and embedding users in the development process prevent good investments from sitting unused.
What leaders should take away
Two main takeaways: first, the technology is significant and leaders must embrace it; second, the CEO’s role is vital in framing AI adoption as an enterprise‑wide business transformation. Swan predicts that in five years the durability of these changes will be clearer: companies that act now and learn from early efforts will establish lasting advantage.
Closing notes
Swan also advised parents and leaders that adaptability and resilience matter most for future success; creativity, communication and empathy remain essential human skills. The refreshed Rewired offers new insights into cutting‑edge AI solutions and the economics of digital and AI transformations, but its central message is consistent: AI creates value when it is embedded into people, processes and leadership, not merely installed as technology.
Dan Swan is a senior partner and global leader, Tech & AI, at McKinsey & Company. Katie Drummond is the Global Editorial Director of WIRED. Views expressed by interviewees are their own and do not necessarily reflect the positions of McKinsey & Company.
