Publish or Perish Paradigm

Research is genuinely fascinating if you have a knack for it. I value the contributions it makes to society. However, I find it challenging to fully grasp what the current research process entails for many researchers out there. I have encountered resources citing personal experiences and anecdotes about how research has become stagnant. To some extent, I can understand these perspectives, but I remain cautious, knowing that individual bad experiences do not tell the whole story.

When I took on the role of a researcher, I was predisposed (though I am not sure how this developed) to the idea that my work would heavily rely on others’ contributions. This remains true and is indisputable. What became unclear as I delved deeper into this role was the unexpected “hype” around AI, which catalyzed a pace of progress that no one could have predicted. It is crucial to mention this because I am unsure if researchers working in AI share this sentiment of needing to publish as rapidly as they can to keep up with the breakneck speed of industry research.

Perhaps the pressure stems from something else. The urgency to publish may have grown due to metrics that universities now use to evaluate potential hires. I suspect such metrics have always existed but have become a top priority in the hiring process. It is fair to say that the goal to publish is not inherently negative; ideally, you want research to be disseminated proactively. However, the publication rate becomes so inflated that it leads to the release of partially novel research and creates unwarranted pressure to meet deadlines.

Does the current state of academia prioritize a high publication rate with weaker research over the potential for well-developed papers? It would be simplistic to claim that those publishing at this inflated rate are not scientists, because to have their work accepted, it must meet scholarly standards. It would be interesting to see where most of these papers are being submitted. The type of conference matters, as respected ones like NeurIPS, Elsevier, and ScienceDirect are highly selective in accepting papers.

Although I do not have direct numbers to support this, I have heard from peers that this is true. However, if no conference is off-limits and evidence shows more papers are being accepted, it warrants further investigation. If we saw that many papers in these conferences are relatively weak compared to past submissions, would that be enough to indicate that research quality is declining? It is hard to tell. I could support this evaluation if we saw that a substantial number of papers are underdeveloped or contain weaknesses that could be strengthened. Pointing to a small percentage of papers that raise questions is not sufficient to support the notion of lower-quality papers put out into academia.

All of these considerations do not even touch on the challenge of staying current with the flood of publications. How should a researcher approach the situation where a minimum of 35 papers need to be read to stay up-to-date in a certain discipline? Is that a reasonable requirement? Of course, to maintain relevance, individuals must adapt. But if we paused for a moment and reflected, should we accept this adjustment? The criteria for selecting these papers will inevitably become stricter to avoid overwhelming researchers with the pressure of reading so many! The alternative would be to “read” all of them superficially.

This is where my main concern lies. I would prefer to know 10 papers by heart than to moderately understand the details of 35 papers. It becomes overwhelming to handle and consider when trying to produce new research. Nowadays, papers are often just skimmed and added to a references section. I feel that papers are being overlooked because there is so much more we could explore and learn from if we took the time to read them thoroughly. Sometimes, I have been frustrated by not being able to delve deeply into papers because true science exists within them! Science is meant to be carried out this way; it is not supposed to be a race to publish faster than others.

I now find myself purely enjoying research at conference paper presentations. It is the only time I feel that outside noise cannot intrude. The pressure to publish and meet deadlines stays out, and logically, this makes sense. If you are attending a conference paper presentation, you are likely also a presenter with a paper awaiting publication. But mainly, I am fully immersed in the papers being presented and make notes or questions for myself. When the presentation ends, I have the chance to ask the lead author about their research. I take advantage of this opportunity because I can learn something new and incorporate it into my daily functions as a researcher! It is a beautiful process to experience, and I just wish research could be like this all the time.