Reskill Yesterday

It's 2 AM in Bangalore. In a fluorescent-lit office park, a woman named "Jennifer" is explaining a billing discrepancy to someone in Ohio. Her real name is Priya. She's been Jennifer for six years now, ever since the accent training taught her to flatten her vowels and bury what the industry calls "mother tongue influence."¹

This is the graveyard shift. Not just a schedule, but a cultural bargain struck over three decades: work American hours, adopt American names, solve American problems. In return, India built a middle class. Over 10 million people employed. 56% of the world's business process outsourcing.² The deal worked because humans were cheaper than geography.

The Irony Nobody Saw Coming

Here's what makes an anthropologist wince: the same qualities that made this work outsourceable make it automatable. Rote. Standardized. Scriptable. Sociologist Shehzad Nadeem noted years ago that outsourced work rarely produces upward mobility because most of it is, by design, routine.³

The scripts Jennifer memorized? AI doesn't need to memorize. It generates.

The scripts she memorized are exactly what AI masters fastest.

The Gap Nobody Talks About

The pattern repeats closer to home. AI use in American workplaces has nearly doubled in two years.⁴ But here's the twist buried in the data: leaders use it frequently (33%), while front-line workers barely touch it (9%).⁴ The people whose jobs are most scriptable are the least equipped with the tools that could make them irreplaceable.

I imagine Jennifer has heard about these tools. Maybe seen a demo in a team meeting, watched someone in management pull up a chatbot that could write the same scripts she spent years perfecting. The graveyard shift isn't just in Bangalore anymore.

What Survives the Night

MIT Sloan researchers identified what AI cannot replicate: empathy, judgment, and what they poetically called "the visualization of possibilities beyond reality."⁵ The ability to sense when a frustrated customer needs something the script doesn't cover. The intuition that comes from six years of midnight conversations with strangers.

What AI cannot replicate: empathy, judgment, hope.

The skills that could be scripted could be outsourced. The skills that could be outsourced can now be automated. What's left are the things that never fit the script in the first place.

The Question

Back in Bangalore, the sun is rising as Jennifer's shift ends. She walks past the training room where she learned to be Jennifer. Someone new is in there now, practicing vowels, learning scripts.

The question isn't whether those jobs change. It's whether the humans in them get to change with them.

Your reskilling plan for 2026 should have started yesterday. But today works too.

---

References

Footnotes

1. Nielsen, K.B. (2025). "Mother Tongue Influence and Global English: Creating 'Neutral' Elites in Delhi's Business Processing Outsourcing Industry." American Anthropologist ↩

2. Ken Research. (2025). "India IT Outsourcing Market Report." Ken Research ↩

3. Nadeem, S. (2015). "Indian Arrivistes and Cyber Coolies: Reflections on Global Outsourcing and the Middle Class." Sociology Compass ↩

4. Gallup. (2025). "AI Use at Work Has Nearly Doubled in Two Years." Gallup ↩ ↩²

5. MIT Sloan. (2025). "These Human Capabilities Complement AI's Shortcomings." MIT Sloan ↩