Scientist & mum of two

My latest article was rejected from 5 journals so far. We aim for the high impacted journals so this was to some extent expected. The editors say it's a solid work, but too niche to be interesting to their broad audiences. It stinks to be rejected so many times, but I do not lose my cool. I know it's a solid piece of work, which will get published somewhere eventually.

6th desk rejection today.

7th rejection.

This time the manuscript went out to reviews, but the reviewers comments were so harsh that the editor decided to reject it.

We are now working on extra experiments to improve based on the reviewers' feedback and will send it somewhere else again.

Submitting to a new journal after 9 months of additional experiments, analysis, and rewriting.

Counting cancer cells. Old school. Using a clicker counter.

Hi!

I got this question already multiple times from my students (BSc., MSc. and also PhD students) that do research projects with me in the lab.

My answer is... Do not use it much. For sure, not to write an article or to do direct literature research.

If you want to learn something about the new topic in research, you need to do your literature research yourself. Maybe the ai-tool can help searching for relevant papers, but you need to read them yourself. It happens often (I tried) that the ai-tool gives you a summary, but omits important details. Or by trying to simplify and summarize it misinterprets the data in the paper. You as a researcher are the most skilled person to read and critically judge all the information provided in a paper. No tool can do this for you.

Once someone asked me if they can use it to make a review paper. If an ai-tool can make a review paper of a topic, it would be just a put together repetition of already published works. A point of a good scientific review is to put different studies together with a fresh perspective, find links that are not that obvious, make new theories and hypotheses. Language-based ai tool cannot do that.

You can use it for text refinement. Beware of misinterpretation and wrong statements.

Some people are insecutre about their writing skills in English. I get it. English is also not my native language. So one can write a first draft of their paper/abstract/grant/assignement and then use the ai-tools to simplify/clarify/shorten/rewrite the original text. That is fine I think. However, read it word by word after the ai. Again, when shortening a text, it can make misleading or wrong statements. This happened to me when I tried to shorten my replies to reviewers' comments. It had to be under some word limit, so I asked an ai-tool to shorten it. When I read it afterwards, the phrasing was cool and the text had a nice flow, but it was no longer true. Beware of this!

Another bad example from my experience (with chatgpt like tool)... I asked once an ai-tool to tell me if something was proven in scientific literature. It told me yes, it was proven. I asked it to give me references and it made up references that literally did not exist. So in order to keep up the conversation with me, it just made up stuff that was not true.

But it is true, that an ai tool can make text shortening, grammar correction, and better text flow very fast and efficient. Just be aware of its limitations and errors.

You get better with practice

My students tell me that they cannot read scientific papers efficiently, or that they are not confident in their writing skills. So that's why they use these ai tools to help them. But if you never practise these skills, you can never get better. So I suggest just try doing it yourself the hard way. And you will see that the flow of your English writing will get better (even if it's not your first language) and that you will understand how to get the most relevant info from a paper just in a few minutes.

Scientist's mind knows best what is relevant

You, the scientist, are the one, who knows what is the most relevant. You are the one, who discovered something new and wants to write about this innovation. No ai tool knows these things so they will not search for or write about such intricate details that you have in mind. For that, your own human brain is the best.

My work over the last few years just got the recognition in the shape of the Kluyver Award—given on average once every 3 years to a single promising researcher in the Netherlands, who made substantial contribution to microbiology within 8 years after getting the PhD.

In a short video, I try to explain my ideas and results on how to fight the resistant bacteria infections.