ML PHD without A* Publications [D]

I know top ML PhD admissions are insanely competitive, so I’m trying to figure out if it’s even worth applying or if I should just focus seriously on jobs instead.

For context, I’m doing my MS at a top-15 US university and have been doing ML research for a while. I’m first author on my projects and mostly work independently, with some guidance from my PI. I had a first-author NeurIPS submission rejected, and I currently have another first-author paper submitted to ICLR, but I’m honestly not very confident about it getting in either.

What’s been getting to me is looking at profiles of people who get into top ML PhD programs. So many of them seem to have multiple NeurIPS/ICML/ICLR/CVPR papers before they even apply, sometimes as undergrads. I genuinely don’t understand how people manage to publish that much that early.

A year ago I was much more confident about doing a PhD. After actually going through the research/publication process, I’ve started doubting myself a lot more. Part of me wonders whether this is just normal and research is hard, especially when you’re doing a lot of it independently. But another part of me is starting to think maybe I’m just not good enough to be competitive for the kind of programs I’m aiming for.

I’m okay with continuing at my current university for a PhD, so this isn’t really a “top program or nothing” situation. But I would like to at least have a realistic shot at some of the stronger ML programs/labs.

The bigger issue is that I’m an international student, so I also need to think pretty seriously about jobs. SWE/MLE recruiting is competitive right now, and I don’t want to spend all my time chasing PhD applications and then realize I’m underprepared for recruiting too. Research roles seem even harder to get without a PhD unless you’re an exceptional MS/BS candidate.

So I’m mainly trying to decide how to allocate my time over the next few months.

If I have strong research experience and first-author projects, but no accepted top-conference papers yet, is it still realistically worth applying to top ML PhD programs?

And for people who were in a similar position, did you still apply, or did you decide to focus on industry instead?

I’m not really looking for “you never know unless you try.” I’m more interested in a realistic assessment of whether the application fees and time are worth it given this kind of profile.