Mumbai’s dabbawalas faced extinction in 2006.
Food delivery apps were emerging. GPS tracking was becoming standard. Investors were funding tech-enabled logistics startups promising faster, cheaper, more efficient delivery. Every expert said the 125-year-old lunchbox delivery system—run by semi-literate workers using bicycles and trains—couldn’t possibly compete.
The dabbawalas had a choice: automate everything or disappear.
They chose neither.
Instead, they looked at their system—5,000 workers delivering 200,000 lunches daily with an error rate of one mistake per 16 million deliveries—and asked a different question. Not “What should we automate?” but “What makes us irreplaceable?”
The answer transformed how they approached technology. They added SMS ordering and a website so customers could book easily. They eventually launched an app for payments and tracking. But they kept humans doing the actual delivery. They kept the color-coded coordination system. They kept the decentralized decision-making where workers adapted to Mumbai’s chaos in real-time. They kept the trust-based relationships with families and offices.
Technology handled what it does best—data processing, payment systems, order management. Humans handled what they do best—navigating unpredictable urban terrain, making judgment calls about delays, maintaining relationships that ensure reliability.
Here’s what strikes me about their approach: they didn’t ask “How do we compete with technology?” They asked “How do we use technology to amplify what makes us valuable?” The result: they’re still operating at Six Sigma efficiency while food delivery apps struggle with consistency and customer satisfaction.
I keep seeing the opposite pattern in organizations. When automation threatens existing roles, most people either fight the technology or try to prove they can do what the technology does—just better, faster, more efficiently. Both strategies fail.
Tom in supply chain management faced this exact choice. AI could optimize logistics networks faster than he ever could. He could have defended his role by emphasizing his optimization expertise. Instead, he doubled down on what AI couldn’t do—understanding business strategy, reading political dynamics in vendor relationships, translating algorithmic recommendations into decisions that account for human factors.
The technology made him more valuable, not less, because he stopped competing with it and started complementing it.
Three patterns I notice in roles that become more valuable as technology advances:
They automate the repeatable, emphasize the judgment-based. Dabbawalas automated ordering and payment processing—tasks that follow clear rules. They kept delivery decisions with humans who navigate traffic, weather, train delays, and building access issues that require real-time judgment. Ask yourself: which parts of my role follow predictable patterns, and which require contextual judgment technology can’t replicate?
They use technology to enhance core strengths, not replace them. The dabbawalas’ app doesn’t replace their delivery system—it makes their reliability more accessible to new customers. Their coding system could integrate QR codes without changing the human coordination that makes it work. Technology became a force multiplier for their existing advantages, not a replacement for them.
They focus on what becomes scarce as automation scales. As AI handles more analysis and optimization, human judgment about which problems matter, which solutions fit organizational culture, and which risks are acceptable becomes more valuable—not less. Strategic thinking, relationship management, ethical reasoning: these don’t automate well. The more routine work disappears, the more premium these capabilities command.
The dabbawalas still operate in 2025. The venture-funded delivery startups that were supposed to replace them have mostly failed or struggle with profitability. Harvard Business School still teaches their system as a case study in organizational excellence.
The executives who promoted Tom to VP of Strategic Operations weren’t ignoring the AI revolution. They were recognizing that leading digital transformation requires someone who understands both the technology’s capabilities and the human judgment needed to deploy it effectively.
The future doesn’t belong to people who can do what technology does. It belongs to people who can do what technology can’t—while using technology to amplify their impact.