Physics undergrad from IIT confused about AI research path
Hi everyone,
I’m a final year undergraduate from one of the top IITs in India. I have a physics background and a CG of about 8.7. Over the past few years, I became very interested in AI and machine learning, especially in building models, reading papers, understanding architectures, and trying to replicate research ideas myself.
What attracted me at first was the depth of the field. I enjoyed spending hours with papers, figuring out why an architecture works. However, lately, with the rise of agents and the current boom in AI, I’ve felt a bit conflicted. Much of what used to seem technically challenging now feels more abstract. This shift has affected my motivation more than I anticipated. It has made me want to explore the mathematical and theoretical side more deeply, rather than just focus on building applications.
This reflection also brought research into my thoughts. I truly enjoy diving deep into problems. I can spend an entire day focused on a single idea and still find it rewarding. Because of this, research seems appealing to me. Yet, I also worry that I might be idealizing it. A PhD is a significant commitment. Once you choose that path, it can be difficult to step back due to time and career impacts. That’s why I’m thinking about pursuing a Master’s first.
Currently, I’m torn among three options:
* Doing a Master’s or research program at my current institute
* Applying abroad for MS programs in AI/ML
* Working for a while first and deciding later
Here are some things that confuse me:
* My CG is good but not remarkable for top AI grad programs
* Pursuing a Master’s abroad is very expensive
* The competition in AI feels overwhelming right now
* I’m not fully convinced that a Master’s is even necessary for the kind of work I want to do
Another honest thought I have is that since I completed my undergrad at a strong institute, I’d like the next step to feel like a real upgrade in terms of academics and research.
I would really appreciate hearing from people who were in a similar situation, especially those who transitioned from physics or math into AI research or industry. How did you determine if research was truly for you? Did a Master’s help clarify your path, or did your industry experience provide more insight?