Three months.
That is how long I waited for a decision on my research manuscript.
Then came the email.
A rejection.
No detailed scientific critique. No discussion of the central hypothesis. No engagement with the mathematical framework.
Just a decision.
At first, I was frustrated.
Then I started asking a more interesting question:
What happens to research that doesn't fit neatly inside one academic box?
That question is much bigger than my own paper.
The Problem With Research at the Boundaries
My work explores an idea called Probabilistic Generative Mind (PGM)—an attempt to think about the mind through probability, generative processes, Markov dynamics, cognition, and artificial intelligence.
It sits somewhere between:
- Philosophy of Mind
- Cognitive Science
- Probability Theory
- Computational Neuroscience
- Artificial Intelligence
- Generative Models
And that creates an unusual problem.
Where exactly does such a paper belong?
Is it computer science?
Philosophy?
Cognitive science?
Neuroscience?
AI?
The answer may be:
All of them.
And that is precisely where modern research infrastructure can become uncomfortable.
Academic systems are excellent at organizing disciplines.
But innovation often happens between disciplines.
A Paper Can Be Wrong.
That's Science.
A hypothesis should be challenged.
A mathematical model should be tested.
An argument should be attacked.
A researcher should absolutely be told:
"Your assumptions don't hold."
That is productive.
That is science.
But there is a difference between:
rejecting an idea after examining it
and
rejecting an idea because it doesn't fit comfortably into an existing structure.
The first protects science.
The second can accidentally protect the boundaries of science.
My Experience With arXiv
My experience with arXiv forced me to think about this distinction.
After months of waiting, the manuscript was not accepted.
The frustrating part wasn't simply the rejection.
Rejection is normal.
The difficult part was having very little scientific feedback that could help answer:
What exactly should be improved?
That leaves researchers in an uncomfortable position.
Should I rewrite the mathematics?
Change the category?
Reframe the theoretical contribution?
Add empirical validation?
Change the philosophical foundation?
Or is the problem simply that the work sits between established categories?
I don't pretend to know the answer.
And I don't think arXiv is inherently "bad."
arXiv performs an enormously important function for science.
But my experience exposed something worth discussing:
Our publishing infrastructure was largely designed around disciplines.
The most interesting AI research may increasingly be designed around problems.
Those are not always the same thing.
Then I Looked Toward PhilPapers
Instead of treating the rejection as the end of the road, I changed strategy.
I began looking at PhilPapers as another route for making the work discoverable by researchers interested in philosophy of mind, cognitive science, AI, consciousness, and related fields.
This isn't about declaring one platform better than another.
It is about understanding something fundamental:
Different research communities have different intellectual maps.
A computer scientist may look at the mathematical structure.
A philosopher may examine the conceptual foundations.
A cognitive scientist may ask whether the model explains cognition.
An AI researcher may ask whether the framework produces computationally testable predictions.
The same paper can therefore look completely different depending on who reads it.
And perhaps that is the real challenge.
AI Is Making This Problem Bigger
Artificial intelligence is accelerating interdisciplinary research faster than our academic categories are changing.
Consider today's AI questions:
Can intelligence emerge from probabilistic generative processes?
That's computer science.
What does it mean for a system to have a model of itself?
That's philosophy of mind.
Can cognition be understood as predictive inference?
That's neuroscience and cognitive science.
Can these ideas be formalized mathematically?
That's mathematics.
Can we build an artificial system based on the theory?
That's AI engineering.
One question.
Five disciplines.
One research problem.
This is becoming normal.
The Next Scientific Infrastructure Should Be Problem-Centric
Imagine a research platform where you don't begin with:
"Which discipline does your paper belong to?"
Instead, you begin with:
"Which problem are you trying to solve?"
A paper about consciousness could automatically connect to:
- generative models
- active inference
- computational neuroscience
- philosophy of mind
- artificial intelligence
- information theory
- cognitive architectures
The platform would create an intellectual graph, not merely a folder hierarchy.
That could change how scientific ideas travel.
Instead of:
Paper → Category → Community
we could have:
Problem → Concepts → Methods → Researchers → Experiments → New Problems
That is much closer to how innovation actually happens.
The AI Researcher of the Future May Not Have a Discipline
This is the part I find most exciting.
Imagine a researcher who develops a theory of consciousness using:
Python.
Bayesian inference.
Markov chains.
Neuroscience.
Philosophy.
Large language models.
Information theory.
And computational experiments.
What should we call this person?
Computer scientist?
Philosopher?
Neuroscientist?
AI researcher?
Maybe the correct answer is:
none of the above.
Maybe the next generation of researchers will be defined less by their department and more by the problems they solve.
Rejection Is Not the Opposite of Progress
My paper being rejected wasn't necessarily a failure.
It forced me to reconsider the path.
Instead of asking:
"How do I get this paper accepted?"
I started asking:
"How do I get this idea in front of the right minds?"
Those are fundamentally different questions.
The first is about publication.
The second is about knowledge propagation.
And perhaps the second question matters more.
A Challenge to AI Researchers
If you are working at the intersection of AI, cognition, philosophy, neuroscience, mathematics, or consciousness, I would genuinely like to hear your experience.
Have you ever developed an idea that didn't fit neatly into an existing academic category?
Did you struggle to find the right venue?
Did reviewers understand the interdisciplinary nature of the work?
Or did the system force you to reshape the idea simply to make it fit an existing taxonomy?
Maybe this isn't an individual problem.
Maybe it is an infrastructure problem.
From Rejection to Experiment
I'm choosing to treat this experience as an experiment.
The manuscript will continue moving.
The ideas will continue evolving.
And the discussion will continue.
Because scientific progress doesn't happen when every paper is accepted.
It happens when ideas survive enough criticism to become stronger.
So perhaps the real lesson from three months of waiting isn't:
"My paper was rejected."
It is:
"The system rejected my submission. It did not necessarily reject the question."
And that distinction matters.
Because sometimes the most important thing a researcher can do after hearing "No" is not to stop.
It is to find a better way to ask:
"Why?"
What do you think?
Should scientific publishing remain primarily discipline-centric, or should we start building problem-centric research infrastructure for the age of AI?
I'm particularly interested in hearing from researchers working across AI + philosophy + cognitive science.
Let's discuss.
created by Seyed Alireza Alhosseini Almodarresieh
Top comments (0)