I’m going to lose trust in science! And that is sorrowful.

In the last few days, I came across two articles, which shocked me in the first place. They also made me sad because it shows how easily and silently the trust in science erodes.

The first case was a video on 🌐YouTube and a 🌐text about the same topic. It compared images from different publications that feature neurons. The creator of the video highlighted that there are suspicious repeating patterns in the pictures, looking like the excessive use of the GIMP stamp tool. What makes me even more worried about that story is that there was an external lab service involved, so even the authors might have been fooled.

The second case was the 🌐report from the editor-in-chief of the Transactions on Machine Learning. He is dealing with an increased amount of poor-quality and AI-“assisted” papers. Nowadays they desk-reject an incredible amount of 53 % of the papers.

He selected 10 of the otherwise desk-rejected papers and gave the authors a fair chance. They should just explain the content of their work, which sounds like a simple task, especially for a single-author paper. Spoiler: It did not go well in most of the cases.

How and why should society trust and believe in science if these practices are out there? How can I have the trust that anything that is written in a paper is honnest work? 

Because for me, science is great, and it is great to learn new things about nature. That is why I wake up each day and go to work. And I know my students, supervisors, collaborators, and friends in person. They share the same spirit, and therefore, I trust them and their work. I can meet them in the lab, and they are happy to show me what they are doing. If I’m lucky, I can even touch the experiment and challenge it.

But how can we build a net of trust on a large scale across science and make sure that honesty wins against shortcuts and frauds?