Two days ago, the headlines screamed a warning that would have made any Indian IT professional's heart sink. Cognition, the maker of the autonomous AI software engineer Devin, had raised a staggering $1 billion at a $26 billion valuation – tripling its valuation in just eight months. More alarmingly, the company revealed an astonishing statistic: 89% of the code committed by its own engineers is now written by Devin.
For millions watching from Bengaluru, Hyderabad, and Pune, this felt like the final verdict. The robot coder wasn't coming; it had already taken over the office.
But then, on May 29, Cognition CEO Scott Wu sat down with TechCrunch and said something that didn't fit the narrative. When asked directly if Devin could replace a mid‑level programmer, he paused. "Yes, and no," he replied. Then he added a line that cut through the panic: "We've never thought about it as replacing humans... it's never been our view."
So, who is telling the truth – the viral LinkedIn posts about the end of coding, or the billionaire founder who built the code‑writing robot? This article goes beyond the headlines to uncover the real future of software engineering in India.
Read also: This $25 Billion AI Coder Just Signed a Deal With Infosys.
The Man Who Codes for Joy: Scott Wu's Philosophy
To understand Devin, you have to understand its creator. Scott Wu is not a business school graduate who stumbled into AI. He is a former child prodigy, a second‑grader who won a nationwide math competition meant for seventh‑graders, which launched a childhood filled with math and programming tournaments. He is an International Olympiad in Informatics gold medalist and has been called one of the most accomplished child competitive programmers of all time.
Wu didn't build Devin to destroy the craft he loves. He built it because he hates "toil" – the tedious maintenance, the painful platform migrations, and the legacy code refactoring that steals the joy from software creation. "We are all programmers ourselves," he told TechCrunch. "I started coding when I was nine."
Read also: Your Manager's Job Is the Next Target: What ClickUp's 22% Layoff Teaches Indian IT
He keeps a physical symbol on his desk: a stuffed teddy bear holding a computer, his own Devin mascot. "This is my buddy who helps you build more," he explained. Wu’s fear is not that AI will take jobs, but that it will take the joy of building away from people. "It's not a secret, most software engineers love building software, right? They love that they can take an idea and turn it into a product."
He wants Devin to act as a new layer of abstraction – just like high‑level languages did for assembly code – freeing programmers to focus on creativity and complex problem‑solving, not repetitive syntax.
Read also: Micron's ₹100 Lakh Crore Milestone: How a Gujarat Plant Powered a Trillion-Dollar Dream
The Indian Reality Check: Infosys and the 'Yes, and No' Paradox
Despite Wu's idealism, the "Yes, and no" reality is already playing out on Indian soil. In January 2026, Infosys announced a sweeping partnership with Cognition to deploy Devin across its internal teams and client projects, starting with the high‑stakes Financial Services practice. The move sent shockwaves through India's tech community, especially among early‑career engineers.
Critics argue that Infosys is replacing the bottom rungs of the corporate ladder. "Infosys would rather pay for Cognition than increase the base pay of junior employees," one user posted on social media. They point to the fundamental business shift: Indian IT services firms are moving from selling "people hours" to selling "outcomes". If an AI agent can fix a bug in minutes without escalation, the need for a bench of thousands of fresh graduates to handle those tasks evaporates.
Infosys itself is also nudging its managers to embed AI more deeply into delivery, with a focus on unlocking higher volumes of work from existing clients. Industry experts warn that as AI agents become more proficient, the demand for "bench" talent and large‑scale campus recruitment may diminish significantly.
Read also: Your Manager's Job Is the Next Target: What ClickUp's 22% Layoff Teaches Indian IT
The Rise of Agentic AI: Markets Don't Care About Sentiments
Scott Wu's philosophy is noble, but the market has its own gravity. The global AI coding tools market, valued at $7.4 billion in 2025, is projected to explode to $24 billion by 2030. This is the fastest‑growing segment in enterprise software, with agentic AI expected to account for 31% of the total generative AI market by 2030, up from just 6% in 2025.
In this gold rush, investors are betting on efficiency, not philosophy. Cognition's revenue run rate has already rocketed to $492 million. Other Indian IT giants are not far behind; TCS and Wipro are exploring similar agentic frameworks to automate maintenance and testing, while Nandan Nilekani has declared that the future is about "shifting from coding to orchestrating intelligence".
The economic pressure is immense. As day‑to‑day coding is commoditised, India’s $200 billion services export engine faces a crisis: global clients won’t pay premium rates for syntax generation any longer; they will only pay for those who can audit intelligence.
Read also: SolarSquare's $60 Million War Chest and the June 1 Deadline That Could Break Indian Rooftop Solar
What Must You Do? The Survival Guide for Indian IT
So, where does this leave the millions of engineers caught in the middle of a Kumbh Milaap (grand confluence) of utopian philosophy and market reality? Here is your roadmap to survive the "Yes, and no" era.
1. Stop Being a 'Syntax Monkey' If your daily work involves writing the same CRUD operations or fixing simple bugs, Devin will do it cheaper. Move up the stack. Infosys leaders are already shifting roles from pure coding to "contextual engineering," focusing on security, architecture, and system integration.
2. Become an 'AI Auditor' Anthropic is actively hiring engineers in India who can "guide, evaluate, deploy, and scale" AI systems, not just write them. The new high‑value skill is not writing code for an AI, but reviewing and verifying the massive amounts of code generated by the AI. Learn to catch hallucinations, security flaws, and logic errors in AI outputs.
3. Embrace the 'Buddy' System Cognition's own story proves that engineers who master agentic tools win. At Cognition, Devin writes the code, but human engineers still architect the system and review the final product. Learn to use tools like Devin, Copilot, or Windsurf as your "buddy." Your efficiency will skyrocket, and you will be entrusted with more complex, creative tasks.
4. Specialise in the 'Messy' Middle AI is excellent at "greenfield" development (starting from scratch). It struggles with "brownfield" chaos – migrating 20‑year‑old COBOL systems or integrating disparate legacy APIs. Becoming an expert in modernisation and complex integration is a career‑proof strategy.
Read also: Crypto's Midlife Crisis: From Lamborghinis to Ledgers, the Indian Reckoning Has Arrived
The Bottom Line
Scott Wu is likely sincere when he says he doesn't want to replace humans. He created Devin to eliminate the "toil" of coding, not the coder. But the Indian IT industry is a ruthless engine of global capitalism. If a tool can increase a developer's output by 5x, the market will not pay for 5 people to do the work of 1.
The true Kumbh Milaap is this: the engineers who survive will be the ones who accept Devin as their "buddy," use it to automate the boring stuff, and then leverage the saved time to solve the complex, architectural problems that no AI can yet understand. The work is not going away. It is just getting much, much harder.
Read also: Meta's 'Real Answers From Real People' Promise Has a ₹13 Lakh Problem
FAQ
Q: What is Cognition's Devin?
A: Devin is an autonomous AI software agent that can plan, write, test, and deploy code end‑to‑end. Unlike autocomplete tools like Copilot, Devin can own entire tasks, such as migrating legacy systems or refactoring codebases.
Q: Is Infosys really using Devin?
A: Yes. Infosys announced a partnership to deploy Devin across its internal teams and client projects, starting with Financial Services. The goal is to boost productivity and modernise legacy systems.
Q: Should Indian freshers stop learning to code?
A: No, but they need to learn differently. The demand for pure syntax knowledge is dropping. Freshers must focus on AI literacy, system design, and cybersecurity to remain employable.
Q: How much is the AI coding market worth?
A: The global market is expected to grow from $7.4 billion in 2025 to approximately $24 billion by 2030, driven by enterprise adoption of agentic AI tools.
Q: Does Scott Wu really believe AI won't replace humans?
A: Wu is a lifelong programmer who built Devin to handle "toil" and maintenance tasks that developers dislike. He envisions a future of human‑AI collaboration, but he admits that Devin is already handling the vast majority of coding work at Cognition.
Read also: AI ka Recharge Bhool Gaye? Why Your Claude Bill Just Went From ₹800 to ₹80,000
Have you started using AI coding agents at your workplace? Do you see them as a threat or a "buddy"? Share your experience in the comments below.
If you found this article useful, share it with a colleague in IT. The era of agentic AI is not coming; it is already here, and the only way to win is to learn to code with the bots, not against them.
Tags: Cognition AI, Devin, Scott Wu, AI Coding Agents, Indian IT Jobs, Future of Work, Agentic AI
Have a question about AI or the latest tech trends? We’d love to hear your thoughts!
Please stay on topic and keep it helpful. Note: All comments are moderated to keep our community spam-free.