Thanks to Andy Yonelinas, ⁦bbuchsbaum⁩ ⁦Dr. Peggy L. St. Jacques⁩ and ⁦Wilma Bainbridge⁩ for a fun symposium. The verdict is in: subjectivity matters! #CNS2024

All “qualitative” judgments are implicitly quantitative.

Causality puzzle: I have a factor and reaction time response. If I vary the factor, observed brain activity differences are “causal”. But what about the factor*RT interaction (difference of slopes)?

GPT and its brethren can’t be tested because they have all the answer keys. They are the most prolific cheaters in history. Obviously, GPT et al. can only pass one test to prove itself: a test of knowledge humans do not yet possess.

This is interesting: GPT4 performs the n-back similarly to humans (would have liked to see a human "control group", though). ojs.aaai.org/index.php/AAAI…

Many historical scientific debates are about continuous vs discrete dichotomies: waves vs particles (physics), symbols vs vectors (AI), slots vs resources (working memory), and, here: brain areas vs gradients.

Also, gradients based on the graph laplacian eigenspectrum are smooth by design. Regions are in some sense the opposite, non-smooth and disjoint.

If Python turns out to be the "language of (AI) thought", a lot of Haskell programmers will be scandalized and indignant.

158 participant fMRI lifespan dataset from Cheryl Grady laboratory at OpenNeuro: openneuro.org/datasets/ds005…

To keep or not to keep, that is the question.

Shortest personality test in the world: the 5 second camera solo each athlete gets as they are introduced to tens of millions before their event.

Are grant reviewers really impressed by a claim of "cutting-edge" methods/techniques?

Using notebookLM a podcast about this paper was created. Enjoy! soundcloud.com/bradley-buchsb…

Looking forward to hosting the 2025 CIHR Summer Program in Aging at Baycrest next June! #SPA2025

Playing around with NotebookLM, challenging them with famously dense and impenetrable works. These facsimile podcasters are completely unfazed and will breeziliy deliver on any topic you throw at them, including Wittengstein's Tractatus soundcloud.com/bradley-buchsb…

If you would like to use the protocol that we used here and in our previous publications (e.g. Olsen, Yeung et al., Neurobiology of Aging, 2017), We have made the atlas available on Dryad! datadryad.org/stash/dataset/…

We anticipate that our paper on quality control (QC) will be an incredibly helpful resource for those of you doing any type of MRI segmentation work. For those doing hippocampal subfield work, we cover common segmentation errors and offer practical guidance on the topic of QC.