Thanks, really good question. Verification and discovery are separate costs. You still have to check every interpretation, that doesn't go away. But the slow bit is usually surfacing relevant work across subfields, not the checking. AI speeds up that search, so your checking time is better spent.

That's some of what I've been up to. My contract's up in October 2026 & I'm starting to look around, so if any of this is useful to you, whether that's a talk, a collaboration, or just a chat, my inbox is open. Always happy to nerd out about faces, methods, or why none of us can agree on a pipeline

Last one, and the most fun: what big team science actually teaches you. We surveyed the people who lead these huge collaborations about everything that goes wrong, funding, diffused responsibility, the lead team quietly walking away, and turned it into practical fixes!

doi.org/10.1525/coll...

And the science itself shouldn't cost the earth. " #Sustainable neuroscience through open science" makes the case that a single #MRI study can run to nearly a tonne of CO₂, and that sharing data openly is how you get more science per gram of carbon 🌍

doi.org/10.1038/s415...

Same instinct, newer problem 🤖 how do you use generative #AI in research without cutting corners? 16 of us ran a #Delphi consensus and landed on eight principles, from data security to accountability, with a checklist to match.

doi.org/10.1007/s436...

#ARTEM-IS is a free webapp that helps you document your EEG/ERP methods properly. You fill in a branching form and it gives you back a tidy PDF and a machine-readable JSON, so someone can rebuild exactly what you did. Loved presenting it at #OHBM #Brainhack2026 this year. doi.org/10.1111/psyp...

Flip the face upside down and the effect disappears, which tells us it's the visual system reading features out in the periphery against a stored template, not just eyes being dragged to the brightest patch. Quietly pleased with that one.

Start with the face. Show someone a fearful or happy face for as little as 50ms and their eyes still get tugged towards the features that carry the emotion, eyes for fear, mouth for happiness. Funny thing is, it doesn't actually help them recognise the emotion any better.

doi.org/10.1038/s415...

The third ongoing project is the perspective paper on using genAI in Biopsychology (led by @annesaulin.bsky.social @ufangyang.bsky.social @michaelgaebler.com @ocklenburg.bsky.social)

Also stay tuned soon for the manuscript! 📣

The second project is the Umbrella Review of EEG Preprocessing for ERP studies (led by @ufangyang.bsky.social) which is currently in the data extraction phase

Great materials! Thanks for sharing this

Shoutout to the amazing team behind it @helenahartmann.com @susannevogel.bsky.social @caggursoy.bsky.social
@alexandroskas.bsky.social @mfjsperl.bsky.social @alinakoppold.bsky.social
Marie Mueckstein, Alexander Lischke @gordonfeld.bsky.social
Grateful for the collaboration & brains on this 💛

A recent MIT study arxiv.org/abs/2506.08872 found that repeated use of AI writing assistants can subtly shift how people reason and write. So how will researchers’ brains change in the long run?
Anyone else curious how neuropsych will tackle this?