Sign in

Aaron Hertzmann

@aaronhertzmann.com
3.8K followers 285 following 409 posts

www.dgp.toronto.edu/~hertzman

PostsRepliesMedia
Aaron Hertzmann @aaronhertzmann.com · 29/09/2026
This paper, its commentaries, and the author response are together the best thing I’ve read in a really long time. Not easy reading, but an amazing resource for debates on consciousness, the nature of the brain, and rejoinders to the “consciousness must be computational” views I hear so often.
141
Reposted by Aaron Hertzmann
Transactions on Machine Learning Research @tmlrorg.bsky.social · 16/09/2026
TMLR has faced a deluge of submissions, necessitating stricter desk rejection policies due to limited reviewer capacity Co-EiC Nihar Shah reached out to authors of 10 papers slated for desk reject. Could they answer questions about their *own* submission? medium.com/@TmlrOrg/ask...
medium.com
Asking Authors About Their Own Papers
By Nihar B. Shah
116163
Reposted by Aaron Hertzmann
Hadas Weiss @hadas.bsky.social · 15/09/2026
sorry about the bad post my account went rogue
1221
Reposted by Aaron Hertzmann
Alison Gopnik @alisongopnik.bsky.social · 01/09/2026
Thanks, but actually my favorite line didn't make it to the piece. Does Odysseus feel guilty about the Trojan war? Lots of discourse on this question even though he doesn't exist! Is he conscious? AI agents are fictive characters abstracted from text - like Odysseus.
28918
Aaron Hertzmann @aaronhertzmann.com · 06/08/2026
I wrote a little piece about how science fiction in movies and TV doesn't care much about accuracy or predicting the future, which is why it's not a good guide for understanding real technology. 1/ aaronhertzmann.com/2026/08/05/s...
aaronhertzmann.com
Science Fiction AI Has Nothing to Do with Real-World “AI”
Science fiction is fiction, not science
3245
Reposted by Aaron Hertzmann
Adverb @adverb.bsky.social · 06/08/2026
machines became superhuman at chess just long enough ago for people to somehow not make any connection between that and every event in 2026
082
Reposted by Aaron Hertzmann
Clément Canonne @ccanonne.github.io · 03/08/2026
FWIW, here are Terence Tao’s slides at the ICM on maths and AI:  teorth.github.io/tao-web/slid... @teorth.bsky.social (I am not endorsing nor criticizing the content, but this is a useful and thoughtful set of points and views in the discussion, from someone who has deeply engaged with the matter)
4347
Reposted by Aaron Hertzmann
Matthew Cobb @matthewcobb.bsky.social · 02/08/2026
Nothing changes.
2105
Reposted by Aaron Hertzmann
Nicholdav @nicholdav.bsky.social · 31/07/2026
hello everyone, I made this helpful meme to explain how to read scientific papers
Vince McMahon reaction meme but its in reverse order from most to least excited and the text to the left of the Vince McMahon faces says title abstract introduction methods results discussion
210724
Reposted by Aaron Hertzmann
Freya Holmér @freya.bsky.social · 22/07/2026
movement in non-euclidean quaternionic space
923931851
Aaron Hertzmann @aaronhertzmann.com · 19/07/2026
This afternoon at #SIGGRAPH2026! I'll be speaking about the human vision in how people draw pictures; excited to see the other speakers!
0100
Reposted by Aaron Hertzmann
Alison Gopnik @alisongopnik.bsky.social · 15/07/2026
This is very good...
091
Aaron Hertzmann @aaronhertzmann.com · 15/07/2026
Storytelling is, first and foremost, about people, and AI in fiction are typically modified people, not genuine explorations of a scientific concept. Viewers can identify with Data as a smart-but-naive nerd, with Terminator (in T2) as an emotionally-closed-off father figure, etc.
050
Aaron Hertzmann @aaronhertzmann.com · 14/07/2026
For years, I've heard from my fellow computer scientists that "the brain is a computer" or "we can simulate brains in a computer." I think both are wrong, and, now that we're hearing them from corporate tech leaders too, it's no longer just an academic topic. 🧵 aaronhertzmann.com/2026/07/13/b...
aaronhertzmann.com
The Brain Is Not Just A Computer, and Cannot Be Simulated by One
Digital computers perform fixed logical operations on clock cycles; brains are biological systems of staggering complexity.
7379
Reposted by Aaron Hertzmann
Carl T. Bergstrom @carlbergstrom.com · 04/07/2026
1. A bit of evolutionary biology. I'm really intrigued by a new perspective piece from Steve Frank that explores connections between how natural selection creates systems that generalize and recent work in machine learning about the surprising capabilities of massively overparameterized systems.
academic.oup.com
Generalization as the great leap in evolvability: insights from machine learning
Abstract. Natural selection encodes learned information in the genome. Learned solutions may be tuned specifically to past challenges, failing in altered e
9402119
Aaron Hertzmann @aaronhertzmann.com · 01/07/2026
On the last day of SIGGRAPH this year, we will hold a new technical workshop called "From Human Vision to Machine Vision", with talks ranging from low-level human visual perception, to applied perception, art & photography, and to machine vision. #SIGGRAPH For more info, see: fromhv2mv.github.io
092
Reposted by Aaron Hertzmann
meetar @fjord.style · 26/06/2026
having “The Talk” with my kid (What Is It Like to Be a Bat? Nagel, 1974)
1506
Aaron Hertzmann @aaronhertzmann.com · 16/06/2026
I haven’t yet finished reading this review, but it already lead me to reading Rosa Cao’s 2022 paper on “multiple realizability” today and it’s my new favorite paper on computational functionalism. Might post most about the key points
260
Reposted by Aaron Hertzmann
yaelvinker.bsky.social @yaelvinker.bsky.social · 12/06/2026
SIGGRAPH 2026 is just around the corner! Planning to attend? Join us for Lines & Minds: Visual Abstraction in Art, Psychology, and Computer Graphics 🎨🧠🫖 We’ll explore how visual abstraction shapes how we think, create, and communicate. 🔗 lines-and-minds.github.io 📅 Sunday, July 19
183
Reposted by Aaron Hertzmann
David Mimno @dmimno.bsky.social · 10/06/2026
This is fascinating work but also I love these happy little circle guys
0274
Aaron Hertzmann @aaronhertzmann.com · 10/06/2026
This piece brilliantly explains why it's not only unnecessary to take seriously the idea that LLMs are conscious, in some cases it is ethically irresponsible.
171
Reposted by Aaron Hertzmann
Alison Gopnik @alisongopnik.bsky.social · 10/06/2026
An excellent thoughtful and clear review of the state of AI. yalereview.org/article/mela...
yalereview.org
Melanie Mitchell: The Dangerous Unknowns at the Heart of LLMs
Melanie Mitchell probes the jagged landscape of AI and its uncertain future.
0285
Aaron Hertzmann @aaronhertzmann.com · 03/06/2026
New blog post pointing out some interesting observations about sketching and how it relates to theories of visual perception. In short, I don't think there's a good account for how scribbly art is recognized (or if there is I don't know it). 1/ aaronhertzmann.com/2026/06/01/s...
aaronhertzmann.com
Scribbling, abstract art, and visual recognition
Sometimes when I sketch quickly, I feel like I’m scribbling some random squiggles, and, yet, recognizable shapes of my subjects emerge:
191
Aaron Hertzmann @aaronhertzmann.com · 29/05/2026
I enjoyed my second #VSS2026 a few weeks ago. I found the session on peer review fascinating, especially the way an editor argued strongly against double-blind review using strawman arguments. Basically “optional blind review doesn’t fully solve all bias, so there’s no point to blind reviewing.”
040
Aaron Hertzmann @aaronhertzmann.com · 26/05/2026
A blog post where I answer the question "prove to me that you are conscious", which AI enthusiasts sometimes use to make AI conscious seems just as plausible as human consciousness. 1/ aaronhertzmann.com/2026/05/20/c...
aaronhertzmann.com
People and Animals are Conscious
“AI” enthusiasts arguing for “AI” consciousness say: prove to me that you are conscious. Of course, one cannot prove this. So this is taken as some kind of “gotcha,” proof of an unfair double standard...
2120
Aaron Hertzmann @aaronhertzmann.com · 19/05/2026
For #VSS2026 attendees: I've posted some of my iPad sketches from the conference on the Discord (in "#chatroom") and here: www.instagram.com/aaronhertzma... A few samples here:
0110
Aaron Hertzmann @aaronhertzmann.com · 18/05/2026
New blog post on dog intelligence, based on my experiences with my own dog and volunteering at a shelter. We often describe dog behaviors in terms of knowledge: “She knows that if she sits, she gets a treat.” “He knows his name" etc. But I think dogs don't have knowledge, they have behaviors. 1/
aaronhertzmann.com
Dogs Do Not Know Things
When you spend time with a dog, you have some ideas about what’s going on in its head. You get a sense of its moods, whether it is happy or tired or afraid. Taking care of a dog, or training it requir...
330
Reposted by Aaron Hertzmann
Christian Wolf @chriswolfvision.bsky.social · 14/05/2026
The Nobel prize in trolling goes to .... somebody who showed a picture of real Monet, pretended it is AI generated (and prompted to imitate Claude Monet), and asked people how they knew it is inferior. People gave detailed descriptions of how its style is wrong and bad. Priceless.
36014
Reposted by Aaron Hertzmann
Shahan Ali Memon @shahanmemon.bsky.social · 30/04/2026
Our position paper, "AI Welfare Is Bullshit" just got accepted to ICML 2026 @icmlconf.bsky.social! The AI welfare agenda has already begun to attract institutional investment at organizations such an @anthropic.com. We argue that this idea is essentially Frankfurtian Bullshit.
56219
Reposted by Aaron Hertzmann
Luba Elliott @elluba.bsky.social · 17/04/2026
My next #LDNCreativeAI meetup is on 13th May with 🖼️ @aaronhertzmann.com @adobe.com on Per-Fixation Perspective in Art and Photography ♟️Vivek Veeriah GoogleDeepMind on Generating Creative Chess Puzzles 🎥 Gabriela Tropia on AI, film and digital necromancy Sign up: luma.com/gx7c3x0i #creativeAI
022
Reposted by Aaron Hertzmann
Paul Graf @paulginva.bsky.social · 18/02/2026
Writing forces your brain to coordinate memory, reasoning, and meaning-making simultaneously. Every time you write, you rewire toward clearer thinking. Every time you let an LLM do it, you rewire toward consumption.
This editorial discusses the critical value of human-generated scientific writing in the era of large language models (LLMs), arguing that writing is essential to structured thinking and research comprehension. 
Writing as Thinking: The act of writing structure's thoughts, sorting research data, and identifying the main message, unlike LLMs which may lack true understanding or accountability.
LLM Hallucinations: LLM-generated text requires rigorous verification because these models can produce incorrect information or fake references.
Human vs. AI Roles: While LLMs are useful tools for brainstorming, improving grammar, or overcoming writer's block, human researchers must maintain control to engage in the creative task of shaping a compelling narrative.
521081
Aaron Hertzmann @aaronhertzmann.com · 26/03/2026
Our internship project earned an Emmy adobe.ly/4rO38ik
adobe.ly
Adobe takes home an Emmy for pioneering work in VR editing tools
Adobe Researchers have been honored with a Technology and Engineering Emmy for their groundbreaking work on virtual reality (VR) editing tools. The Emmy celebrates Adobe’s “significant contributions to the development of 360-degree consumer video capture, editing, and presentation technologies.”
2120
Aaron Hertzmann @aaronhertzmann.com · 25/03/2026
If asked to draw a person's portrait, many people would say “I cannot draw.” Why is observational drawing so hard? In a new paper, @judithfan.bsky.social and I answer: it's the limitations of human vision. Learning to draw is learning skills to overcome them. aaronhertzmann.com/2026/03/23/d... 1/
aaronhertzmann.com
Why Drawing is Hard: Visual Limitations and the Skills to Overcome Them
If asked to draw a picture of a tree or a person in front of them, many people would say “I cannot draw.” Thirty years ago, two psychologists pointed out that this should be surprising. They reasoned ...
210635
Aaron Hertzmann @aaronhertzmann.com · 22/02/2026
If navigation apps like Google Maps were invented today, they would be called "AI" and hyped as "AI agents." It's truly amazing how well they work, and now everyone completely takes them for granted, and many people completely outsource the mental task of navigation to them.
090
Reposted by Aaron Hertzmann
samuel mehr @mehr.nz · 08/02/2026
free video for intro lectures on auditory perception
412726
Aaron Hertzmann @aaronhertzmann.com · 02/02/2026
Considering a big change to some slides, I find the Mitch Hedberg line to be apt: “I write jokes for a living, I sit in my hotel, I think of something that's funny, then I go get a pen and I write it down. Or if the pen is too far away, I have to convince myself that what I thought of ain't funny.”
020
Reposted by Aaron Hertzmann
Clément Canonne @ccanonne.github.io · 21/01/2026
What is wild to me is the defense, BY THE NEURIPS BOARD, that fabricated citations do not mean "the content of the papers themselves [is] necessarily invalidated" It does. It very much does. What do you think citing other work is for? What do you think writing a paper is for? What do you *think*?
1624775
Reposted by Aaron Hertzmann
Jessica Hullman @jessicahullman.bsky.social · 09/01/2026
Wrote a summary of a great keynote by @zey.bsky.social at NeurIPS, arguing that we’re having the wrong nightmares about AI: not AGI or superhuman benchmarks, but good-enough genAI at scale threatens "load bearing frictions" society relies on to signal effort, authenticity, sincerity, credibility.
open.substack.com
Zeynep Tufecki on having the wrong nightmares about generative AI
I was writing a blog post where I was going to reference Zeynep Tufecki’s 2025 NeurIPS keynote, and realized there isn’t a solid synopsis online.
28027
Reposted by Aaron Hertzmann
derek guy @dieworkwear.bsky.social · 07/01/2026
A Washington Post article suggests that, beyond a certain threshold, larger homes don't make people happier. Instead, well-being is correlated with affordable housing in walkable neighborhoods where they feel socially connected. www.washingtonpost.com/climate-envi...
Excerpt reads: "Prioritize your neighborhood
Neighborhoods drive our happiness (or unhappiness) more than we think.

When Lina Martinez, director of POLIS, the Center for Wellbeing Studies at Universidad Icesi in Colombia, studied the happiness of households in the city of Cali, she couldn’t find much difference between poorer and richer areas. “Their happiness is pretty much the same,” she said. But that changed when she isolated neighborhood conditions, such as access to transit, health care and parks. “The conditions of the neighborhood affect happiness,” she said. “I can’t link that to the space where [people] live.”

That aligns with findings from a 2023 study of the Vancouver metro area: Researchers found no significant differences in well-being between people living in single detached homes, duplexes, townhouses, laneway houses and apartment buildings (basement units smaller than 300 square feet were the only negative exception). What did people say they missed most in their neighborhood? Affordability, proximity to family and friends, and a sense of community. Home design was eighth on the list."
19471021003
Reposted by Aaron Hertzmann
William Ngiam | 严祥全 @williamngiam.github.io · 31/12/2025
Hey everyone! I have enough publications to stop working for the rest of the year!
3281
Aaron Hertzmann @aaronhertzmann.com · 22/12/2025
It’s tough to listen to recordings of my dog barking and howling intermittently while she learns to stay home alone, but I did like this auto-transcription produced by the audio recording app (Apple Voice Memos)
180
Reposted by Aaron Hertzmann
Angelina Wang @angelinawang.bsky.social · 12/12/2025
Most LLM evals use API calls or offline inference, testing models in a memory-less silo. Our new Patterns paper shows this misses how LLMs actually behave in real user interfaces, where personalization and interaction history shape responses: arxiv.org/abs/2509.19364
13811
Aaron Hertzmann @aaronhertzmann.com · 12/12/2025
People say that the new "AI"-based ads look like garbage, but the Brian Eno quote is relevant: people learn to appreciate the quirks of a style. The tastes here are about associations, e.g., if people who hate "AI" take the glitches as signifying "AI", then people will hate the glitches.
280
Reposted by Aaron Hertzmann
Ethan Mollick @emollick.bsky.social · 09/12/2025
Historian Thomas Hughes argued that technologies are malleable when young, then harden. Right now we're still shaping AI, or at least it is being shaped by our institutions, norms & use cases Eventually these systems build a momentum of their own. That is why choices now matter, things are fluid.
3559
Aaron Hertzmann @aaronhertzmann.com · 02/12/2025
I’m really enjoying @wdavidmarx.bsky.social 's wonderful new book “Blank Space.” He argues that 21st century “poptimism,” a belief in the value of all styles of music has led to a loss of cultural innovation; we no longer have movements of disco rejecting rock, punk rejecting disco, etc. 1/
370
Aaron Hertzmann @aaronhertzmann.com · 26/11/2025
As we in the US gather with our families this weekend, it's a good time to consider the ways that technology brings us together—but also keeps us apart. In this blog post, I tell a story about how technology may have tended to increase social isolation. aaronhertzmann.com/2025/10/26/i...
aaronhertzmann.com
Technology and Social Isolation: From Cars to “AI”
In this blog post, I tell a story of how some technologies from the past century or so have, overall, led to increased social isolation in the United States, as we replaced in-person social interactio...
074
Reposted by Aaron Hertzmann
Molleindustria FUTURE? NO THANKS! OUT NOW @molleindustria.org · 26/11/2025
Today I learned that the US poverty line is still calculated as 3x the minimum food diet in 1963, a metric completely disconnected from today’s cost of living. Here's a clear post explaining why middle and working class families are effectively getting poorer. www.yesigiveafig.com/p/part-1-my-...
yesigiveafig.com
Part 1: My Life Is a Lie
How a Broken Benchmark Quietly Broke America
1276
Aaron Hertzmann @aaronhertzmann.com · 12/11/2025
These technologies illustrate how much of our culture follows common patterns
040
Reposted by Aaron Hertzmann
Mark Riedl @markriedl.bsky.social · 07/11/2025
This is the best parody of the AGI2027 report to date. Unintentionally from the Dallas Fed
1274