Will Harrison
LightNews LightNews
banner
willjharrison.bsky.social
Will Harrison
@willjharrison.bsky.social
890 followers 390 following 600 posts
Vision scientists and horror film enthusiast.

I also lecture at the University of the Sunshine Coast when I'm not at the beach.
Posts Replies Media Videos
Reposted by Will Harrison
racheldenison.bsky.social
Happy to share this new paper from the lab led by Angus Chapman, now out in PLoS Biology! It presents an integrated spatiotemporal normalization model for continuous vision. @afchapman.bsky.social

journals.plos.org/plosbiology/...
A dynamic spatiotemporal normalization model captures perceptual and neural effects of spatial and temporal context
The effects of spatial and temporal context on sensory systems have mostly been studied independently of each other. This study shows that the modulation of visual perception and neural activity by th...
journals.plos.org
December 18, 2025 at 6:30 PM
willjharrison.bsky.social
The idea to use a bot (probably) to write about the idea that we could be running out of new ideas is itself a new idea, thereby proving that we are not, in fact, running out of new ideas.
jasonsanford.bsky.social Jason Sanford @jasonsanford.bsky.social · 16d
This column on Vox is the biggest pile of steaming AI crap I've seen this year. Human's aren't running out of ideas. Instead, those with power and money don't want to listen to ideas that threaten their power and money.

Ideas are plentiful. Putting those ideas into action is the hard part.
We’re running out of good ideas. AI might be how we find new ones.
What if the best use of AI is restarting the world’s idea machine?
www.vox.com
December 14, 2025 at 2:11 AM
Reposted by Will Harrison
There is money just sitting on the table for a university brave enough to grab it.
victorerikray.bsky.social Victor Ray @victorerikray.bsky.social · 24d
The first school to market itself as AI free is going to corner the market on people interested in actually learning. And I would not be surprised if rich families and the children of people creating this tech were the first movers.
jeffsharlet.bsky.social Jeff Sharlet @jeffsharlet.bsky.social · 24d
My employer, Dartmouth College, today boasts it's 1st Ivy "to launch AI at an institutional scale." It is doing this by partnering--"more than a collaboration"--with Anthropic, a company that stole the books of many faculty, me included, which many of us are suing.
December 5, 2025 at 4:20 AM
Reposted by Will Harrison
williamngiam.github.io
Do you have an open working memory dataset and want it to be findable and reused? You can now add it to the Open WM Data Hub: williamngiam.github.io/OpenWMData! The collection of datasets tagged with useful metadata is steadily growing thanks to a small team of volunteers!
OpenWMData
A collection of publicly available working memory datasets
williamngiam.github.io
December 1, 2025 at 11:28 PM
Reposted by Will Harrison
jephpp.bsky.social
Thank you to the 2025 IJE cohort who helped improve manuscripts this year via our pre-external review process!

Matthew O'Donohue: scholar.google.com.au/citations?us...
Marcell Székely: scholar.google.com/citations?hl...
Zekun Sun: zekun-sun.github.io
December 3, 2025 at 11:48 PM
Reposted by Will Harrison
kostaboskovic.bsky.social
New pre-print with @drbarner.bsky.social! We ask how children come to understand age. We find that young children use numerical age and facial morphology to identify who’s older, not just size, and point to acquiring a number system as key to developing an understanding of age.
osf.io/gvb46
OSF
osf.io
December 1, 2025 at 5:04 PM
Reposted by Will Harrison
ruthrosenholtz.bsky.social
My Visual Attention in Crisis paper has finally appeared, along with 30 commentaries and response. I argue that it’s time to rethink attention from the ground up, and suggest key phenomena and possible directions. Requires access, alas. doi.org/10.1017/S014...
Visual Attention in Crisis | Behavioral and Brain Sciences | Cambridge Core
Visual Attention in Crisis - Volume 48
doi.org
November 26, 2025 at 1:23 PM
willjharrison.bsky.social
Delicious.
November 26, 2025 at 1:26 AM
Reposted by Will Harrison
hakwan.bsky.social
despite the questionable choice of keynote speaker, i think this is likely going to be a good conference
sampendu.bsky.social Sam Schwarzkopf @sampendu.bsky.social · Nov 23
It is our great honour to announce the APCV Keynote at #epc-apcv-2026: Prof. Hakwan Lau from the Institute for Basic Science in Korea!
The call for member-initiated symposia & abstracts is now open: visualneuroscience.auckland.ac.nz/epc-apcv-2026

#psychscisky #visionscience #neuroskyence
November 25, 2025 at 1:04 AM
Reposted by Will Harrison
irisvanrooij.bsky.social
See also for instance: irisvanrooijcogsci.com/2025/08/12/a...
AI slop and the destruction of knowledge
Cite as: van Rooij, I. (2025) AI slop and the destruction of knowledge. This week I was looking for info on what cognitive scientists mean when they speak of ‘domain-general’ cognition. I was curio…
irisvanrooijcogsci.com
November 23, 2025 at 4:21 PM
willjharrison.bsky.social
One of the few papers of my own that I enjoy reading is an analysis of a wonderful image database. But I'm a nerd.

Harrison, W. J. (2022). Luminance and Contrast of Images in the THINGS Database: Perception, 51(4), 244–262. doi.org/10.1177/0301...
Contrast Energy The contrast energy of THINGS images was well described as a linear transform of spatial fre- quency on log-log axes, with a slope of 1.3 (Figure 2A). There was relatively little variation across images, with the simple 1/f model accounting for 97% of variance across all 26,107 images. Contrast energy as a function of orientation, however, was relatively more heterogeneously distributed across images (Figure 2B). Nonetheless, our model of oriented contrast accounts for 17% of the variance in the THINGS images. There are many plausible reasons for the greater con- sistency of energy as a function of spatial frequency relative to orientation. As one example, Torralba and Oliva (2003) showed that such statistics depend on whether image content is natural or not, with relatively less cardinal-oblique contrast bias for natural images. A strength of the THINGS database is its diversity of image content. Furthermore, by design, the images in the THINGS database are digital photos with objects in various compositions and configurations. For example, some photos tightly frame a specific object, and others are framed with the camera lens rotated by some amount. These variations will have little impact on the 1/f spectrum of the images, which simply quantifies the scale invariance of the contrast distributions, but can greatly impact the distribution of oriented contrast (e.g. a picket fence rotated by 45° will have the same 1/f spectrum as the original, but its oriented contrast distribution will be translated). It is somewhat unsurprising, therefore, that oriented contrast is more variable across images. Perhaps more import- ant to the intended use of THINGS images, however, is that there is no reason to expect that these distributions of contrast energy deviate meaningfully from those encountered in natural settings.
November 21, 2025 at 9:14 AM
Reposted by Will Harrison
martinhebart.bsky.social
Please repost! I am looking for a PhD candidate in the area of Computational Cognitive Neuroscience to start in early 2026.

The position is funded as part of the Excellence Cluster "The Adaptive Mind" at @jlugiessen.bsky.social.

Please apply here until Nov 25:
www.uni-giessen.de/de/ueber-uns...
View from your office onto Giessen and surrounding villages.
November 4, 2025 at 1:57 PM
Reposted by Will Harrison
meganakpeters.bsky.social
This is wild!
initself.bsky.social Victoria Bosch @initself.bsky.social · Nov 3
Introducing CorText: a framework that fuses brain data directly into a large language model, allowing for interactive neural readout using natural language.

tl;dr: you can now chat with a brain scan 🧠💬

1/n
November 3, 2025 at 6:12 PM
Reposted by Will Harrison
silverpebble2.bsky.social
A reminder that the original Hallowe'en jack o' lanterns date back to 18th century Ireland if not earlier, were based on a shady boozy blacksmith called Stingy Jack who cheated the devil & was trapped between 2 worlds, were carved from turnips and looked like this:
A preserved carved turnip jack o'lantern from the 19th century A preserved carved turnip jack o'lantern from the 19th century
October 31, 2025 at 7:59 AM
willjharrison.bsky.social
Great thread and I wanna throw in my two cents too: increasing N does not increase the chance of a false positive per se (theoretically it'll always be alpha).

However, increasing N, particularly with lots of DVs, makes very small confounds more likely to produce a statistically significant result.
annemscheel.bsky.social Anne Scheel @annemscheel.bsky.social · Oct 31
There still seems to be a lot of confusion about significance testing in psych. No, p-values *don’t* become useless at large N. This flawed point also used to be framed as "too much power". But power isn't the problem – it's 1) unbalanced error rates and 2) the (lack of a) SESOI. 1/ >
November 1, 2025 at 12:56 AM
Reposted by Will Harrison
amanitatermy.bsky.social
*I* would never say this of course, but SOME people might say it's almost like university managements are in on the grift.
disabilitystor1.bsky.social Aparna Nair @disabilitystor1.bsky.social · Oct 28
Maybe the thing that gives me a pulse-pounding throbbing headache is when I think of how techbros sold ChatGPT to universities as ‘revolutionizing’ higher education and our institutions just forked over money without even asking for actual evidence
drjoehanson.bsky.social Joe Hanson @drjoehanson.bsky.social · Oct 28
When people learn with ChatGPT instead of following their own searches, they end up knowing less, caring less, and producing worse advice, even when the facts are the same.

Friction is an essential ingredient for learning! Convenience makes us shallow.

academic.oup.com/pnasnexus/ar...
October 28, 2025 at 9:09 PM
willjharrison.bsky.social
Huge career highlight for me: I got to chat about the psychological science of horror movies with legendary actor John Jarratt, and about how he created his iconic role in Wolf Creek.
#RecreationalFear

Article written by Thomas Fowles:
www.usc.edu.au/about/unisc-...
Holy s**t, something bad is about to happen! The art vs the science of scary cinema
Dr Will Harrison and actor John Jarratt explore the psychological tricks used in Australian horror films Picnic at Hanging Rock and Wolf Creek, and why audiences want to be thrilled.
www.usc.edu.au
October 27, 2025 at 1:02 AM
Reposted by Will Harrison
emilya-izzeddin.bsky.social
As this is my final PhD paper to be published, I want to give a special shout-out to my supervisors, @willjharrison.bsky.social and Jason. There’s no way to do them justice in 300 characters, so I’ve attached an excerpt from my thesis acknowledgements that still rings true today.

11/11
October 8, 2025 at 7:13 AM
willjharrison.bsky.social
The real pain won't be felt for another decade, not that I'm convinced the US will still exist then.
October 14, 2025 at 1:07 AM
willjharrison.bsky.social
New paper looking at how manipulating a movie's static statistics can influence subsequent perceptual inference.
emilya-izzeddin.bsky.social Emily A-Izzeddin @emilya-izzeddin.bsky.social · Sep 29
Thanks for reading if you've made it this far! I've had to skip a lot of the details, so if you're interested in learning more, feel free to have a read of the paper - here's the link again:

www.nature.com/articles/s41...

10/10
Investigating orientation adaptation following naturalistic film viewing - Scientific Reports
Scientific Reports - Investigating orientation adaptation following naturalistic film viewing
www.nature.com
September 29, 2025 at 11:19 PM
willjharrison.bsky.social
We got people to watch Casablanca after we manipulated the movie's image statistics in one of four ways. Emily came up with a cool way to probe people's orientation estimation during the movie. Paper now out - details in the thread 💪
emilya-izzeddin.bsky.social Emily A-Izzeddin @emilya-izzeddin.bsky.social · Sep 29
The movie itself was filtered frame by frame to have contrast at a specified adaptor orientation at low spatial frequencies. We completed this process four times, so we could test different adaptor orientations (0, 45, 90, and 135).

2/10
September 29, 2025 at 10:03 PM
willjharrison.bsky.social
Punny footnote:

"It is not entirely clear what prompted the cow connection. Informally we referred to this project as the multi-site overview of oscillations (MOO) which got the ball rolling, after which we decided to milk it for all it was worth."

royalsocietypublishing.org/doi/10.1098/...
How strong is the rhythm of perception? A registered replication of Hickok et al. (2015) | Royal Society Open Science
Our ability to predict upcoming events is a fundamental component of human cognition. One way in which we do so is by exploiting temporal regularities in sensory signals: the ticking of a clock, falli...
royalsocietypublishing.org
September 18, 2025 at 1:17 AM
willjharrison.bsky.social
I'm only just learning about splitapply in matlab:
au.mathworks.com/help/matlab/...
au.mathworks.com
September 16, 2025 at 11:05 PM
Reposted by Will Harrison
olivia.science
Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia:
doi.org/10.5281/zeno...

We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
Abstract: Under the banner of progress, products have been uncritically adopted or even imposed on users — in past centuries with tobacco and combustion engines, and in the 21st with social media. For these collective blunders, we now regret our involvement or apathy as scientists, and society struggles to put the genie back in the bottle. Currently, we are similarly entangled with artificial intelligence (AI) technology. For example, software updates are rolled out seamlessly and non-consensually, Microsoft Office is bundled with chatbots, and we, our students, and our employers have had no say, as it is not considered a valid position to reject AI technologies in our teaching and research. This is why in June 2025, we co-authored an Open Letter calling on our employers to reverse and rethink their stance on uncritically adopting AI technologies. In this position piece, we expound on why universities must take their role seriously toa) counter the technology industry’s marketing, hype, and harm; and to b) safeguard higher education, critical thinking, expertise, academic freedom, and scientific integrity. We include pointers to relevant work to further inform our colleagues. Figure 1. A cartoon set theoretic view on various terms (see Table 1) used when discussing the superset AI (black outline, hatched background): LLMs are in orange; ANNs are in magenta; generative models are in blue; and finally, chatbots are in green. Where these intersect, the colours reflect that, e.g. generative adversarial network (GAN) and Boltzmann machine (BM) models are in the purple subset because they are both generative and ANNs. In the case of proprietary closed source models, e.g. OpenAI’s ChatGPT and Apple’s Siri, we cannot verify their implementation and so academics can only make educated guesses (cf. Dingemanse 2025). Undefined terms used above: BERT (Devlin et al. 2019); AlexNet (Krizhevsky et al. 2017); A.L.I.C.E. (Wallace 2009); ELIZA (Weizenbaum 1966); Jabberwacky (Twist 2003); linear discriminant analysis (LDA); quadratic discriminant analysis (QDA). Table 1. Below some of the typical terminological disarray is untangled. Importantly, none of these terms are orthogonal nor do they exclusively pick out the types of products we may wish to critique or proscribe. Protecting the Ecosystem of Human Knowledge: Five Principles
September 6, 2025 at 8:13 AM
willjharrison.bsky.social
Why everyone needs to understand signal detection theory (or at least sensitivity versus specificity).
September 3, 2025 at 12:03 AM