In 2025 I realized how good the AI-assisted coding tools are becoming, and so I start experimenting to try and understand the technology better, and to see whether I can leverage this technology to do more good and boost science. With the launch of Claude Code and Codex CLIs and the jump in abilities with models like Opus 4.5 and GPT5 AI in late 2025, I felt like I’m gaining super-powers, and that I can now do things I never imagined possible before for a single researcher.

Since then I have built many tools aimed at better understanding and boosting science.
You can see an overview of the tools in Scirenity. I aim to build an infrastructure of tools and platforms that would allow any researchers to do better more transparent and credible science.

I am now giving talks and workshops to try and show others how to use these superpowers:

If you’re in need of a talk or a workshop, or are doing something related that you think we can work on together, then please do reach out.

Disclosures

Lots of what I now do is AI created or AI assisted.

My profile photo, many of the recent presentations, posters, and thumbnails, and ofcourse many of the tools and guides I make available, are built and/or generated with the assistance of AI.

It would seem that some people hold strong views against such usage, and I think I understand hesitation, confusion, caution, and resistance. I would welcome having a conversation on that, if you’re interested.

Way I see it: I aim to learn by doing, to use this technology and improve use, see how people use my tools and react to what I do. There is no ignoring the fact that people are increasingly using these, and so, as part of the process, it is for us to experiment with these.

As far as I’m concerned, I support and encourage the use of AI, as long as things are not harmful to anyone, and are transparent, clear, and openly disclosed.