Vinod Kumar, from PwC's manufacturing practice, was on a podcast recently. One point he made deserves its own conversation.
India's manufacturing is at 16% of GDP. The target is 25%. The national ambition is 25%. Vinod Kumar is clear that this gap cannot be closed by copying solutions built for other countries.
His argument was about redirection, not shortage.
India already has the talent. The question is whether it gets applied here.
On what AI actually changes inside a factory, he was specific:
📍 Digital twins of equipment and processes: not dashboards, but working replicas that give operators real-time insight into throughput, cost, quality, and reliability. Decisions that used to live in a senior engineer's head now have data underneath them.
📍 Support processes running on agentic AI: routine work handled without human intervention, so people move toward higher-judgment tasks instead of administration.
📍 Demand planning using unstructured and image data: signals that traditional planning systems never captured, now feeding into pricing and inventory decisions before the market moves.
But the conversation that actually matters happens before any of this.
On ROI, Vinod's point was direct on how AI should not be treated as an operating expense. It is a capital investment, the same way you would budget for a new factory or mine. The bar should be just as high.
That one reframe changes what gets approved and what doesn't.
The talent is here. The technology exists. What's missing in most mid-market companies is the boardroom perspective to treat this as infrastructure, instead of experimentation.
How is your leadership team currently thinking about AI investment - expense or capital commitment?