Twitxy Saved

Sai Dwivedi

Verified account This account is verified. Learn more

@saidwivedi

PhD Candidate at @MPI_IS | ex: @Meta, @MercedesBenz, @intel
Tübingen, Deutschland saidwivedi.in Joined July 2011
971 Following 743 Followers 211 Posts

Latest posts 5

Sai Dwivedi Verified account @saidwivedi
“It never pays to be religious about technology.”

I’ve made this mistake quite a few times in the past, where technology becomes your identity. Now I always remind myself that technology is a tool to solve a problem.
Michael Black Verified account @Michael_J_Black Nov 6
The dogma of the last decade in computer vision is that all classical algorithms should be replaced by neural networks. This has been very productive but, counterintuitively, with the latest frontier models, I find myself sometimes turning back to classical methods.

If you have the training data, the dogma still holds. But there are situations where data is hard to get, there exist good classical algorithms for solving a problem, optimization methods are just better, or where some simple code just does the thing you want.

When writing code was time consuming and error prone, training everything made sense. Now, that a frontier model can code up and test solutions quickly, getting a solution through coding can be faster than gathering training data and training a model. Consequently, I think there will be more hybrid approaches that blend trained models and code.

We now have generic, high-level, planners that can execute sequences of lower-level tasks to solve large problems. These lower-level tasks may be solved by a neural network, a classical method, or specialized code written on the fly.

It never pays to be religious about technology. The future is probably hybrid. Pick the thing that works for you.
View quoted post on X
Sai Dwivedi Verified account @saidwivedi
I find it difficult to understand the mindset of deliberately breaking an already flawed system, hoping something better will emerge.

I agree that academic peer review is broken. I hate reviewing AI slop just as much as I hate receiving AI-generated reviews of my own papers.

But why are people doing this? Because the system rewards publication counts. It’s also easy to blame the system, but when 500 people apply for one research position, you need some way to filter candidates. Publications aren’t a great metric, but we still haven’t found a better way to evaluate research at that scale.

Deliberately flooding conferences with low-quality papers, hoping the system will collapse, only makes things worse for everyone trying to do meaningful research.
Michael Trinh Verified account @_mctrinh Oct 12
AI researchers are destroying themselves.
View quoted post on X
Sai Dwivedi Verified account @saidwivedi
Before debating whether a PhD is worth it, I think it’s important to ask: What is the purpose of a PhD?

Historically, it has served different purposes:
— Before 1100 AD: In early Christianity, the title doctor was associated with teachers of theology.
— 1100–1800 AD: Universities awarded doctorates to certify scholars as qualified to teach.
— 1800 AD onward: PhDs increasingly focused on producing new knowledge through original research.

Today, one of the main goals of a PhD today is to train independent researchers. But the opportunities to develop those skills have changed, especially in today’s era.

Research papers and open-source code are more accessible than ever. Of course, becoming an independent researcher still requires good mentorship, collaborators, compute, and the freedom to explore difficult problems. A PhD can provide that environment, but so can a strong industry research lab.

For me, pursuing a PhD made sense. I had spent four years working in industry in India, where getting access to the kind of frontier research environment I wanted was difficult. A PhD opened that door, and I’m glad I went for it.

A PhD was the right choice for me, but that doesn’t mean it’s the right choice for everyone. Before committing several years to a PhD, I think it’s worth asking: What do I actually want to get out of it, and is it the best way to achieve that?
MIT CSAIL Verified account @MIT_CSAIL Oct 10
Wise words from Geoffrey Hinton: A PhD in computer science is worth it.

v/@samuelxzhang
View quoted post on X
0 replies 8 reposts 84 likes 17K views View on X
Sai Dwivedi Verified account @saidwivedi
Many PhD students say they lose control when vibe coding. But I think of it this way: if Michael can usually figure out what went wrong with my project just by looking at my presentation, without ever looking at my code, I should be able to do the same with my coding agents.
Michael Black Verified account @Michael_J_Black Oct 10
PhD Advising vs Vibe Coding #1:
One of the key skills a PhD advisor learns is context switching. Students come into your office every 30-60 minutes and you need to remember their project, where you left off, and what they were supposed to do next. It's a learned skill and exactly what you need to manage AI agents.
View quoted post on X