(BPT) - Why the incoming college class must trade control for oversight to thrive in the age of AI.
By Arvind Nithrakashyap, CTO and Co-Founder, Rubrik
As you step onto campus this fall, the prevailing narrative about the job market likely feels like a ghost story. You've heard that AI is coming for entry-level roles, that coding is dead and that the degree you're working toward might be obsolete before you cross the stage at graduation.
As a parent of a teen coming of age, and co-founder of a tech startup that went public in the past two years, I'm here to tell you that the reality is far more interesting, and the opportunity is far more compelling.
College students should not see AI as a threat to the jobs they hope to hold. Rather, you should view AI as an opportunity to enter the workforce already knowing and understanding how to direct, evaluate and improve the work of AI agents. Technology skills are no longer reserved for those pursuing tech degrees. Professionals across every industry will need to build them, and now is your time to rise to the challenge.
If you listen to the "AI-washers" in Silicon Valley, you'd think every company is already a perfectly oiled AI machine. The "flying cars" of AI haven't arrived yet because the bottleneck isn't the technology; it's the people. To succeed in this world post-graduation, you don't need to fear the bot. Your advantage will come from combining human judgment with the ability to direct the bot effectively.
From doer to manager of agents
Today's college students are essentially the first generation to enter the workforce already prepared to lead with AI. You have an advantage; you do not have to unlearn decades of workplace habits that your predecessors did prior to AI.
The most profound shift you can make during your college years is adopting a change management mindset. Historically, an individual contributor was expected to be a doer who personally touched every line of code or every page of a legal brief.
In the agentic era, you must learn to operate like a manager.
Across the industry, we are seeing people use AI to move beyond the traditional boundaries of their jobs. Meanwhile, some historically technical roles have struggled because they refuse to let go of direct control. They feel they must personally review every step an agent takes, moving only at the speed of a human rather than the speed of a machine.
To succeed, you must get comfortable with acceptable risk. You have to be willing to let an agent run for hours, what I call "YOLO (you only look once) mode," and audit the results after the fact rather than micromanaging the process.
The proof is in the transformation
All of this is not just theory. At Rubrik, we are deep in the throes of our own AI transformation. We're seeing these mindset shifts drive meaningful results across every department. For example:
- Legal: Our data privacy legal team, without any engineering help, built an agent that reads and analyzes product code for privacy and data governance gaps. Using this tool moved the team from a block-and-review model to an inform-and-audit model, allowing innovation to move fast while they oversee the stream of activity and maintain portfolio-wide activity.
- Product Management: Our PMs are now building functional prototypes to show customers directly, bypassing the engineering backlog entirely by using AI to bridge the technical gap.
- Operations: A single program management leader automated a manual task that used to require every team to build 10–15 slides for quarterly reviews, turning it into a dynamic dashboard.
- Engineering: We've seen roadmap capabilities pulled forward by four months simply because our teams embraced coding agents.
The common thread with these examples is not technical expertise alone. The people creating the greatest value know how to define a problem, give an AI system appropriate boundaries, evaluate its work and remain accountable for the result.
Your mission: Reframe the college experience
During college, the fundamentals of your expertise area are still critical; you still need to learn how to write, analyze, build and reason without relying on AI. Then learn how AI can extend those abilities. Practice breaking large problems into clear assignments. Learn to test outputs, identify weak reasoning and recognize when a decision requires human judgment and taste.
Don't spend the next four years trying to protect your domain from AI. Spend them learning how to govern it. Students who develop foundational expertise and AI leadership skills will not simply be prepared for the future of work. They will help define it.
The job market isn't shrinking; it's evolving. The door is wide open for those who enter their professions not as laborers, but as architects of autonomous systems.
Welcome to the agentic era.
