Pat Savage's Avatar

Pat Savage

@patrickesavage.bsky.social

Director: @compmusiclab.bsky.social. Rutherford Discovery Fellow @U Auckland. Assoc. Prof. @Keio U. PI @manyvoices.bsky.social. Music, evolution, diversity. He/him. Tangata tiriti.

2,509 Followers  |  712 Following  |  715 Posts  |  Joined: 04.09.2023  |  1.9441

Latest posts by patrickesavage.bsky.social on Bluesky

My book has an official website! (Though the full book won't be officially published for a few more weeks)
global.oup.com/academic/pro...

09.02.2026 09:52 β€” πŸ‘ 5    πŸ” 2    πŸ’¬ 0    πŸ“Œ 0
Preview
Visual language models show widespread visual deficits on neuropsychological tests - Nature Machine Intelligence Tangtartharakul and Storrs use standardized neuropsychological tests to compare human visual abilities with those of visual language models (VLMs). They report that while VLMs excel in high-level obje...

Our latest paper, β€œVisual language models show widespread visual deficits on neuropsychological tests”, is now out in Nature Machine Intelligence: www.nature.com/articles/s42...

Non-paywalled version:
arxiv.org/abs/2504.10786

Tweet thread below from first author @genetang.bsky.social...

09.02.2026 02:40 β€” πŸ‘ 37    πŸ” 20    πŸ’¬ 1    πŸ“Œ 1
Post image

Thrilled to share our latest paper, out now in Science Advances! We explored the development of cooperative behaviors β€” fairness, trustworthiness, forgiveness, & honesty β€”Β  across five societies, culturally contextualizing them & seeing how they correlate. (1/5) www.science.org/doi/full/10....

07.02.2026 15:09 β€” πŸ‘ 101    πŸ” 32    πŸ’¬ 1    πŸ“Œ 2
Top left: Experimental paradigm. The authors analyzed EEG data recorded from 49 sleeping human newborns while being exposed to monophonic piano melodies composed by J. S. Bach (real condition) and control stimuli (shuffled condition). Top right: Surprise and entropy. Surprise and entropy associated with each note’s timing (green, St and Et, respectively) and pitch (yellow, Sp and Ep, respectively) were estimated using an unsupervised statistical learning model trained on all stimuli. Dot plots display mean surprise and entropy associated with real and shuffled music, averaged across melodies (left panel), and separately for each melody (right panel).  Bottom: Analytical approach. Multivariate Temporal Response Function (mTRF) models were fit to describe the forward relationship between multiple stimulus features and the EEG signal. The full TRF model (leftmost panel) included acoustic low-level features (spectral flux, acoustic onset, IOI, and IPI) and high-level features (surprise and entropy of pitch and timing).

Top left: Experimental paradigm. The authors analyzed EEG data recorded from 49 sleeping human newborns while being exposed to monophonic piano melodies composed by J. S. Bach (real condition) and control stimuli (shuffled condition). Top right: Surprise and entropy. Surprise and entropy associated with each note’s timing (green, St and Et, respectively) and pitch (yellow, Sp and Ep, respectively) were estimated using an unsupervised statistical learning model trained on all stimuli. Dot plots display mean surprise and entropy associated with real and shuffled music, averaged across melodies (left panel), and separately for each melody (right panel). Bottom: Analytical approach. Multivariate Temporal Response Function (mTRF) models were fit to describe the forward relationship between multiple stimulus features and the EEG signal. The full TRF model (leftmost panel) included acoustic low-level features (spectral flux, acoustic onset, IOI, and IPI) and high-level features (surprise and entropy of pitch and timing).

Does our very human ability to anticipate #musical structure exist at birth? @robertabianco.bsky.social @giacomonovembre.bsky.social &co show that #newborns encode #rhythmic (but not melodic) expectations based on statistical regularities in real #music @plosbiology.org plos.io/4kqKVWg

06.02.2026 13:55 β€” πŸ‘ 22    πŸ” 4    πŸ’¬ 0    πŸ“Œ 1

bsky.app/profile/patr...

07.02.2026 00:20 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Preview
How Jeffrey Epstein Bankrolled An Exclusive Intellectual Boys Club And Reaped The Benefits The Edge Foundation runs what has been called the β€œworld’s smartest website” and held annual β€œbillionaires’ dinners.” It was also financed by Jeffrey Epstein and gave him access to elite circles in sc...

Academics:
If you’re trying to make sense of how people you know knew Epstein, it’s book agent John Brockman. He’s savvy enough to lay low, but we should be talking about him.

Friends don’t let friends sign with Brockman.

www.buzzfeednews.com/article/pete...

And

newrepublic.com/article/1548...

04.02.2026 16:17 β€” πŸ‘ 47    πŸ” 24    πŸ’¬ 4    πŸ“Œ 2
Screen shot of a youtube video titled "AI.FILL Function Explained: 10X Productivity in Excel with AI" with the caption "Let ChatGPT fill your missing data"

Screen shot of a youtube video titled "AI.FILL Function Explained: 10X Productivity in Excel with AI" with the caption "Let ChatGPT fill your missing data"

Don't you f**king dare.

03.02.2026 14:56 β€” πŸ‘ 305    πŸ” 66    πŸ’¬ 28    πŸ“Œ 39

Key to many of these connections was Brockman, "Epstein's intellectual enabler": newrepublic.com/article/1548...

04.02.2026 01:08 β€” πŸ‘ 3    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image

The current reforms of the research system are now on track for another failure. They double down on mistakes that undermine long-term growth, well being & safety. We must ask for accountability and transparency about three key issues as we go into the election year.

scientists.org.nz/news/13593197

02.02.2026 20:31 β€” πŸ‘ 20    πŸ” 16    πŸ’¬ 1    πŸ“Œ 2

Can anyone create a Google Scholar plugin for this?

02.02.2026 09:49 β€” πŸ‘ 15    πŸ” 0    πŸ’¬ 0    πŸ“Œ 2
Post image

Wiley: "We’re supporting responsible research assessment practices" onlinelibrary.wiley.com/journal/1520...

Also Wiley: "Prove that your article is a good fit for this journal πŸ˜‰πŸ˜‰πŸ˜‰πŸ˜‰πŸ˜‰ by citing at least two of our articles in your manuscript before we will even consider reviewing it" 🀑

30.01.2026 11:29 β€” πŸ‘ 65    πŸ” 31    πŸ’¬ 9    πŸ“Œ 8
Preview
Elon Musk Emailed Extensively With Jeffrey Epstein, Begging to Visit His Notorious Island Newly released documents show how Elon Musk extensively communicated with Jeffrey Epstein to plan parties and trips to his island.

NEW: Elon Musk has insisted that Jeffrey Epstein invited him to the notorious island and he declined. But new emails show the opposite -- Musk was practically begging to party with Epstein in the Caribbean, years after Epstein's sex convictions were public knowledge

futurism.com/future-socie...

30.01.2026 20:25 β€” πŸ‘ 5156    πŸ” 2005    πŸ’¬ 99    πŸ“Œ 336

Exciting to see this launch. Been in the works for a long time, with dozens or possibly hundreds of journals involved, and major library institutions, banding together for sustained support of free to read, free to publish in, open access journals. #openaccess

29.01.2026 19:24 β€” πŸ‘ 16    πŸ” 5    πŸ’¬ 0    πŸ“Œ 0
Bilbo looking at his phone top

on bottom is ChatGPT

After all, why not? Why shouldn't I keep it?

You're absolutely right β€” you found it, it's been with you a long while, and it's only natural to feel fond of something that's served you so well, especially when someone like Gandalf suddenly seems to want it for himself.

Bilbo looking at his phone top on bottom is ChatGPT After all, why not? Why shouldn't I keep it? You're absolutely right β€” you found it, it's been with you a long while, and it's only natural to feel fond of something that's served you so well, especially when someone like Gandalf suddenly seems to want it for himself.

28.01.2026 01:51 β€” πŸ‘ 25377    πŸ” 6953    πŸ’¬ 7    πŸ“Œ 155

Evergreen reminder that this was something she came up with, not something she was advised to do. In fact she was pretty specifically advised it was a terrible idea (it was) but she went ahead with it anyway because something something Not Real Science

28.01.2026 01:59 β€” πŸ‘ 129    πŸ” 64    πŸ’¬ 5    πŸ“Œ 2
Preview
β€˜Bizarre’ linguistics paper on water retracted by Springer Nature Jordanian professor’s free-wheeling study is pulled and an editorial board member sacked after linguists criticised β€˜extremely strange’ article

β€˜Bizarre’ linguistics paper on water retracted by Springer Nature www.timeshighereducation.com/news/bizarre...
@lameensouag.bsky.social @dingemansemark.bsky.social @verbingnouns.bsky.social

21.01.2026 08:10 β€” πŸ‘ 24    πŸ” 6    πŸ’¬ 0    πŸ“Œ 2
Preview
Death Bed: The story of Kelly Savage A young New Zealander entered a Japanese hospital in psychosis. He left braindead. His grieving mother is fighting for change.

I'm finding Alex Petti's murder especially hard to process because he's almost the same age my brother Kelly would have been if he had not also been killed. Details are different, but in both, the people in power decided not to investigate themselves.
shorthand.radionz.co.nz/kelly-savage...

26.01.2026 12:02 β€” πŸ‘ 6    πŸ” 3    πŸ’¬ 1    πŸ“Œ 1
Agents appear to shoot at him at least ten times within five seconds, beginning while he was pinned to the ground and continuing after he collapsed and his body lay motionless.

Agents appear to shoot at him at least ten times within five seconds, beginning while he was pinned to the ground and continuing after he collapsed and his body lay motionless.

McDonald was walking away from the police when he was shot 16 times by Officer Van Dyke

McDonald was walking away from the police when he was shot 16 times by Officer Van Dyke

Thinking of this song a lot: "16 SHOTS":
www.youtube.com/watch?v=KPWX...

RIP Alex Pretti.
RIP RenΓ©e Good.
RIP George Floyd.
RIP Laquan McDonald.
RIP USA.

26.01.2026 11:43 β€” πŸ‘ 4    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

(Or are using it to try to cover your tracks…)

26.01.2026 05:03 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

It’s a problem if you’re trying to crank out AI-generated slop without getting caught… πŸ˜₯

26.01.2026 04:26 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0
Dr Kareem Carr
man: i wish to publish
@kareem_carr
Jan 21
reviewer 2: your paper is no good
man: i'll do anything to improve
reviewer 2: it's simple. you must read the work of the great scientist Pagliarini
man: *bursts into tears* but i am Pagliarini
Andre Pagliarini
@apagliar
Jan 21
a first: in rejecting an article I submitted to a journal, reviewer 2 noted I failed to engage the work of one Andre Pagliarini
Jan 21, 2026 β€’ 3:47 PM UTC

Dr Kareem Carr man: i wish to publish @kareem_carr Jan 21 reviewer 2: your paper is no good man: i'll do anything to improve reviewer 2: it's simple. you must read the work of the great scientist Pagliarini man: *bursts into tears* but i am Pagliarini Andre Pagliarini @apagliar Jan 21 a first: in rejecting an article I submitted to a journal, reviewer 2 noted I failed to engage the work of one Andre Pagliarini Jan 21, 2026 β€’ 3:47 PM UTC

I just thought everyone should see this

22.01.2026 23:02 β€” πŸ‘ 25537    πŸ” 6064    πŸ’¬ 43    πŸ“Œ 235
Preview
When two years of academic work vanished with a single click After turning off ChatGPT’s β€˜data consent’ option, Marcel Bucher lost the work behind grant applications, teaching materials and publication drafts. Here’s what happened next.

Amazing sob story: "ChatGPT deleted all the work I hadn't done"

www.nature.com/articles/d41...

22.01.2026 19:51 β€” πŸ‘ 1536    πŸ” 396    πŸ’¬ 120    πŸ“Œ 273
Video thumbnail

Top of the charts at my house today

17.01.2026 14:43 β€” πŸ‘ 1681    πŸ” 542    πŸ’¬ 27    πŸ“Œ 99
11.01.2026 03:41 β€” πŸ‘ 6248    πŸ” 1722    πŸ’¬ 29    πŸ“Œ 33
Preview
How Nature Became a 'Prestige' Journal Since launching in 1869, Nature has evolved from a periodical offering commentary on pigeons to the prestige journal in science. But how did Nature build its reputation, and can it last?

"Ultimately, peer review didn’t transform Nature’s standing so much as protect it, converting exclusivity that might have seemed arbitrary into gatekeeping that appeared meritocratic."

Nice piece by Robert Reason on the prestige of Nature www.asimov.press/p/nature

11.01.2026 06:57 β€” πŸ‘ 25    πŸ” 13    πŸ’¬ 0    πŸ“Œ 0

It is as it ever was. elevanth.org/blog/2022/01...

10.01.2026 09:47 β€” πŸ‘ 21    πŸ” 5    πŸ’¬ 1    πŸ“Œ 0
Jason Gay
@jasongay
There's a guy in this coffee shop sitting at a table, not on his phone, not on a laptop, just drinking coffee, like a psychopath.

Jason Gay @jasongay There's a guy in this coffee shop sitting at a table, not on his phone, not on a laptop, just drinking coffee, like a psychopath.

Me at Midnight Espresso sitting at a table, not on my phone, not on my laptop, just drinking coffee, like a psychopath.
(Well, technically I used my phone to take the picture…)

Me at Midnight Espresso sitting at a table, not on my phone, not on my laptop, just drinking coffee, like a psychopath. (Well, technically I used my phone to take the picture…)

Did this for the first time in ages - felt great, highly recommend

10.01.2026 09:21 β€” πŸ‘ 11    πŸ” 3    πŸ’¬ 0    πŸ“Œ 0

NA->0 strikes again!
@rmcelreath.bsky.social

10.01.2026 09:09 β€” πŸ‘ 12    πŸ” 2    πŸ’¬ 0    πŸ“Œ 0
Preview
Simulation-based inference with deep learning suggests speed climbers combine innovation and copying to improve performance Abstract. In the Olympic sport of speed climbing, athletes compete to reach the top of a 15 m wall as quickly as possible. Since the standardization of the

Proc B with @sampassmore.bsky.social! We used simulations to explore the innovation strategies of speed climbers πŸ§—β€β™€οΈ Innovation is higher among slower athletes and lower when the population size is larger, and the overall balance of innovation and copying appears to be suboptimal πŸ”— bit.ly/499QjZM

08.01.2026 14:00 β€” πŸ‘ 42    πŸ” 18    πŸ’¬ 3    πŸ“Œ 4
A 3D image of a speed climbing wall, next to a graph showing the decline in speed climbing times since 2007.

A 3D image of a speed climbing wall, next to a graph showing the decline in speed climbing times since 2007.

The cultural evolution of Speed Climbing bit.ly/499QjZM

New article w @masonyoungblood.bsky.social

Innovations are more likely to occur in slower athletes & large populations. Simulations suggest innovation is underutilized!

08.01.2026 19:06 β€” πŸ‘ 16    πŸ” 3    πŸ’¬ 1    πŸ“Œ 1

@patrickesavage is following 20 prominent accounts