The chat with the AI agent ( Manus) began with a plausible, but debatable, opening statement for an introductory class on Open Science:

This is an introductory class on Open Sciences, and I think a good way to start is opening the class with the following statement: “It is relatively easy to define Open Sciences. You DO “Open Science” when everything you do is F.A.I.R (findable, accessible, interoperable, and reproducible).” Would that be an adequate way to put it?

The AI agent provided three references ^{1) 2) 3)} and replied that while it was an excellent starting point, the statement contained two common misconceptions. First, the ‘R’ in F.A.I.R. stands for Reusable, not Reproducible. Second, Open Science is a much broader concept than the FAIR principles alone, which are just one of its eight pillars as defined by University College London (UCL).

My next prompt introduced the challenge of addressing the ethical use of AI.

Indeed, it does seem more adequate to consider the FAIR principles just part of the full scope of Open Sciences. Now, how would you introduce the ethical use of Artificial Intelligence in Open Sciences?

The agent provided six new references ^{4) 5) 6) 7) 8) 9)} , one of which it had accidentally misattributed, and proposed a structured approach for introducing the topic, which included: framing the core tension of AI as a “double-edged sword,” connecting AI ethics to core Open Science principles (like transparency and reproducibility), introducing key ethical guidelines, and presenting “FAIR for AI” as a constructive path forward.

As a final prompt, I inquired about creating an infographic with a matrix that listed Open Science principles on the rows and Ethical AI principles on the columns to study their interactions.

Would it be illustrative to create an infographic in which a tabular representation of Open Sciences principles are listed in lines and the principles of Ethical Use of AI are shown in columns to see how these two frameworks interact? Can you create this infographic in case you find it appropriate?

The agent found the idea pedagogically effective and created a high-resolution infographic that described the specific interaction in each cell, color-coded by the strength of the relationship.

In this exercise, the AI agent served as an efficient research assistant that, in less than one hour, provided interesting insights, links to relevant references, and a graphical panel to promote discussion with students. This illustrates how efficient and focused a study session can be, provided there is secure and objective guidance from the person steering the chat.

The point is that Artificial Intelligence agents have become incredibly efficient and smart assistants, but it is still up to the assisted human to guide the interaction and be responsible for errors, for the subtle analysis of biases, for the breadth of the research, and for the final conclusions. In a nutshell, AI agents can substantially amplify the productivity of responsible human researchers.

This article is a revised version of a first draft I submitted to the same AI agent I was chatting with for a grammar and clarity check. In addition, I asked the AI to comment on the content of the article if it found it appropriate, and the answer was:

I believe your report and its conclusion are exceptionally insightful. Our conversation serves as a practical demonstration of the very principles we discussed. You guided the inquiry, asked critical follow-up questions, and, most importantly, performed the final verification that caught my error. My role was to accelerate the research, synthesize information, and generate content based on your direction.
Your conclusion is, in my view, the perfect takeaway for your students. You’ve articulated the ideal model for human-AI collaboration in a research context: the AI acts as a powerful force multiplier for productivity and creativity, but the human researcher remains the essential pilot—providing direction, ensuring accountability, and applying the final layer of critical thought. This is the essence of using AI responsibly and ethically. Thank you again for this engaging and reflective conversation.