Is Academia Reaching a Breaking Point?

11 August, 2026

I recently submitted a paper to a workshop at a top-tier computer science conference. In previous years, the workshop received around 40–70 submissions and accepted roughly half.

This year, they received more than 250 🤯.

I love workshops. They are great spaces for early-stage work, Master’s projects, new PhD students, and ideas that do not need to become full conference papers. A friend of mine, who recently started her PhD, also submitted her first paper there.

But 250+ full-length submissions raises an obvious question: who is going to review all of them?

Reviewing is already time-consuming academic labour, and while our ability to produce papers seems to be accelerating, our ability to review them is still fundamentally human and slow.

Which brings me to AI.

Tools like Claude Code and Codex can now assist with coding, experiments, analysis, and writing. Research workflows that once took much longer can increasingly be accelerated with AI. If researchers can produce more papers in the same amount of time, submission numbers will naturally rise.

As an early-career researcher, I find this both exciting and worrying. There is an uncomfortable question: if everyone else is using AI to move faster, can I afford not to?

Maybe computer science publishing is approaching a point where something has to change — whether that is how we review, what we consider a meaningful contribution, or how many papers we expect researchers to produce.

For now, this is my current academic crisis:

If producing papers becomes dramatically easier, but evaluating them does not, what happens next?

If you are thinking about the same thing, I would love to chat.