Humans as an API call for AI.
Surprise winner of the IIT Alumni Council’s hackathon announced on April 17, 2025
As a shocking and scary demonstration, a Kodoy ai agent put out a free ad in June 2025 on a job portal for hiring therapists to treat its patients of minor Physio therapy issues. And real humans applied and started working. And then those who applied were put on another portal seeking customers. Customers paid the service providers directly.
It was not just a gimmick — it is a preview of how work will function in 2030. And if you can hire a therapist, why not a contract murderor or anything else.
AI is no longer just a tool that humans use. It is becoming a system that hires humans on demand. This seemed far fetched when in the final presentation they showed us a real website doing exactly what was feared.
It has over 4,800 people listed, 134 tasks completed, and payments made in crypto — small numbers today, but a very big signal. Here is the link to those wanting to enrol:
Payments are direct and in crypto. Paid by the AI Agent from its own crypto wallet. No kyc is required for such a wallet.
This flips the traditional relationship between people and technology. Instead of humans directing machines, machines decide what needs to be done and call a human only when reality gets in the way: when a physical presence is required, when a door must be opened, when a face must appear, or when a human judgment call is unavoidable.
In this world, humans are not employees in the classical sense. They are becoming endpoints in a global API — summoned, used, and released when the task is complete. Reliability, availability, and location matter more than résumés.
For decades we feared that AI would replace us. The more uncomfortable truth is different: many of us will not be replaced, we will simply be rented by intelligent systems.
The real dividing line in the future of work will not be skilled versus unskilled.
It will be those who build and control these systems versus those who are plugged into them.
We have millions happily willing to plug into any free AI. Sure it does a good job of search or of writing a report. But what if it kills your first born. Or poisons your whole family to a condition far worse than death.
And we still think AI needs no policing.
Or that rogue technologies are faraway.
What the winner did demonstrate is a technology stack to ensure that a human is not working for a AI agent.