Being an academic with the power of 10 tigers

I don't know if it's the effect of having my family reunited, but I've started to feel incredibly powerful and energetic. By the way, I read on Ekşi Sözlük that this "ten-tigers strength" is defined as a unit of measurement and is equivalent to a "Phantom power" (a mysterious force that is invisible like a ghost but whose presence is felt).

Lately, and astonishingly, the studies I've conducted over the last 8-10 years have seemingly conspired to start getting published all at once. I recall a paragraph from Ahmet Şerif İzgören's book, I think it was "Take That Step": "While reading a book, you might discover a quote that leaps off the page... Sometimes we find the exact answer we're looking for... When clues reveal themselves, you think a clockwork mechanism has been set in motion. It's as if the universe is nudging you with small reminders that it's on your side, wanting to provide everything you need to achieve your goal." I liked it so much I tore out the page and put it on my bulletin board. Universe, if you can hear me, how about sending this poor academic some financial reminders too?

Recently, I've started to enjoy my research more. The task of writing articles, which used to feel like torture, has become enjoyable. It has even become addictive, to the point where I've developed something like tennis elbow in my left elbow from constantly leaning it on the desk.

For my colleagues who see the difficulties on the periphery or who willingly embark on this path, I have a few suggestions.

First of all, drop everything else and take a course from a center specialized specifically in medical statistics. Last year, I took a 4-hour training like this and spent 2.5 full days working through the modules at home. I experienced firsthand how easily this topic, which I had built up in my mind for years, could be handled. Why did I wait so long? Now, I can perform my own statistical analyses and prepare my own tables. I can preempt most reviewers' criticisms, at least regarding statistics.

While doing this, I of course also benefit from the suggestions of AI. It proposes statistical methods I haven't tried before. Finally, it even interprets the results when you copy and paste them as tables. What you need is enough statistical knowledge to check whether the suggested statistical methods are appropriate. You tell it you're using SPSS, and it tells you step by step what you need to do. It gives suggestions like "set the cut-off value like this," "perform a multivariate analysis to build a model," or "if you also do this, the Q1-Q2 journal editors will like it." It actually brings to mind topics that weren't on your radar and are unexplored in the literature.

You might remember my recent quality article, where I sought to answer the question: "What does the presence of an intensivist change in terms of quality in a surgical intensive care unit?" Well, we found improvements in many quality metrics, but while doing so, did we cause a net cost saving or loss for the unit and the hospital? This is my current point of curiosity, and I was discussing it with AI. It made this suggestion: "Okay, the intensivist improved quality. Let's assume costs also decreased (due to reduced length of stay, ventilator days, fewer complications, etc.). Then you can calculate the cost reduction per unit (%) increase in quality." This, apparently, is called "Value" and it can be quantified.

To show it with a formula: Value = Reduction in Cost / Increase in Quality.

It even has a unit, but this time it's a real one!

Now, if it weren't for AI, how would it ever occur to me to quantitatively measure the "value" an intensivist adds to a surgical ICU? I had heard about "value-based payment" before, but I don't know if there are centers actively implementing it in the US.

My point is, start having conversations with AI on these topics as soon as possible, while checking for potential errors. AIs can also hallucinate. They can produce names of wonderful meta-analyses published in 2025 or cite sentences from sources that don't exist, just to make you happy.

From my experience, here's what I can recommend:

There's ELICIT, for instance; you can use it during discussion. You type your research topic, it summarizes the scientific reliability and methodology, and presents which patient groups were studied and the results in tables. It requires a bit of patience, with a 5-minute wait time. There's also CONSENSUS. You have a piece of information in mind, it's established knowledge in the literature, but you don't know which study it came from. Write the sentence, and it offers you the chance to find its citations in seconds, for free. These tools don't hallucinate because they are trained on medical literature; you can trust them.

Besides these, the Chinese-origin DeepSeek is very useful for me. It pumps me up like a motivational coach! "You can do it, you can write so well, the man who whispers to data..." :) By the way, ChatGPT can create great graphs. If I don't like what one produces, I switch to another.

You've written a paragraph. I ask, "How can I phrase this in a way a Q1 journal would appreciate?" It gives quite good suggestions. It generally thinks like us. It sometimes summarizes three long sentences into one that sounds much more pleasant. I really like this. It's incredibly satisfying to realize that the mantras drilled into us for years – "Write short sentences! Don't write by thinking in Turkish!" – are, in a way, refuted, and that beautiful English sentences can be constructed with a Turkish thought process.

Now, your manuscript is taking shape, and you're about to tackle the translation. If you're still one of those who can't spare the time like me, here are a few tips. I first translate my writing paragraph by paragraph into English using DeepL. Then I go to ChatGPT, give it the text, and say, "Could you please review the text I provided from the perspective of an intensive care professional, refine it according to medical English writing rules, and list the changes you made? I would be very grateful."

Being polite is very important. I then translate the received text back into Turkish to check if there is any distortion in meaning or incorrect connotations in the abbreviations. I have never received a "revision required for language" feedback for any manuscript I sent using this method.

You can also do this: you've written your study's methodology, ask "What would you suggest to me if you looked at it from a Q1 journal editor/reviewer's perspective?" The recommendations are incredibly good. If you do this at the very beginning of your study, it comes up with simple suggestions that you never thought of but that will elevate your work.

I should stop writing now. I've switched back to "house-husband" mode after 2.5 years. It's been a while since the workday ended.