iiinova — where your digital self keeps evolving

Aaron Hertzmann @aaronhertzmann.com · 06/08/2026

This suggests a fun exercise: watch fictional sympathetic robots, but imagine them as mere chatbots: Data providing simulated relationships to humans; C-3PO + R2-D2’s bickering as generative performances of personality; WALL-E mindlessly following its programmed behaviors long after humans left. 5/

110

Aaron Hertzmann @aaronhertzmann.com · 04/08/2026

Very interesting! It seems like there's a common trend here; a similar thing happens in visual art when medium-quality models make the most unintentionally-interesting pictures: dl.acm.org/doi/abs/10.1... osf.io/preprints/ps... (artbreeder images are deliberately manipulated for effect)

110

Aaron Hertzmann @aaronhertzmann.com · 14/07/2026

Chaotic systems like double pendulums are especially hard to get right and make accurate predictions. So simulation of even moderate size systems can be nearly insurmountable. We can predict accurate behavior of the weather with some probability, but it's far from a complete simluation. 12/

100

Aaron Hertzmann @aaronhertzmann.com · 14/07/2026

Maybe the brain is some kind of non-digital computer? This might be true, but not useful. We don't have much theory for analog computers beyond outdated technology like slide rules. "It's an analog computer" sounds like "it's a computer." It's really misleading. A brain is just a brain. 8/

Nordsieck's Differential Analyzer from 1950

100

Aaron Hertzmann @aaronhertzmann.com · 14/07/2026

I agree, but I think biology/mechanisms play a role in how we value others. For example, modern values about animal welfare often relate to beliefs about their natural behaviors. My dog doesn't have the right to vote. Anyways, I think it's worth arguing both fronts (social and substrate).

The five freedoms for animals from the Humane Society.

The Five Freedoms ensure that we meet the mental and physical needs of animals in our care:

Freedom from hunger and thirst by ready access to fresh water and diet to maintain health and vigor. This must be specific to the animal. For example, puppies, adult dogs, pregnant cats, and senior cats all need different types of food provided on different schedules.
Freedom from discomfort by providing an appropriate environment including shelter and a comfortable resting area. This means you should provide soft bedding and an area with appropriate temperature, noise levels, and access to natural light. If an animal is outside, it must have shelter from the elements as well as appropriate food and water bowls that will not freeze or tip over.
Freedom from pain, injury, or disease by prevention or rapid diagnosis and treatment. This includes vaccinating animals, monitoring animals, physical health, treating any injuries and providing appropriate medications.
Freedom to express normal behavior by providing sufficient space, proper facilities, and company of the animal's own kind. Animals need to be able to interact with — or avoid — others of their own kind as desired. They must be able to stretch every part of their body (from nose to tail), and run, jump, and play. This can be particularly challenging when animals are housed in individual kennels.
Freedom from fear and distress by ensuring conditions and treatment which avoid mental suffering. The mental health of an animal is just as important as its physical health — as psychological stress can quickly transition into physical illness. These conditions can be achieved by preventing overcrowding and providing sufficient enrichment and safe hiding spaces.

110

Aaron Hertzmann @aaronhertzmann.com · 14/07/2026

It comes down to the point from your paper about different usages of "computer". When people say that LLMs can be conscious because the brain is computer, that isn't about analog computers.

120

Aaron Hertzmann @aaronhertzmann.com · 14/07/2026

FYI I am a computer scientist and I taught undergrad and grad courses in computer science for many years. I have worked through the steps of Turing machines by hand. I understand the difference. That's not what the debate here is about.

110

Aaron Hertzmann @aaronhertzmann.com · 14/07/2026

By "computer," we normally mean the real, digital computers that convert everything to 0s and 1s and process them with logic circuits, all synced to a fixed clock cycle. You can separate what a computer computes from how: the same software can run on many kinds of computers. 3/

230

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

Aaron Hertzmann @aaronhertzmann.com · 03/06/2026

And here's a progression of Piet Mondrian's work over a decade before he got to his most famous abstract work. 8/

100

Aaron Hertzmann @aaronhertzmann.com · 03/06/2026

Some scientists have claimed that the aesthetics of abstract art relate to real-world natural image statistics, which I agree with. And I think this is visible in the way that some abstract artists began realistic and got more and more abstract; here's a progression of Richard Diebenkorn's work. 7/

110

Aaron Hertzmann @aaronhertzmann.com · 03/06/2026

Here are strokes optimized to be classified as faces, in 2020, by @mmariansky.bsky.social 6/

100

Aaron Hertzmann @aaronhertzmann.com · 03/06/2026

As a model for this, consider the work of @drib.net , in which he optimized strokes and blobs to maximize state-of-the-art image classifiers (in 2018) to produce desired categories. This is his "ceiling fan." medium.com/artists-and-... 5/

110

Aaron Hertzmann @aaronhertzmann.com · 03/06/2026

Even when there are recognizable outlines, a lot of the detail seem like textural scribbles, not recognizable individual details. 3/

100

Aaron Hertzmann @aaronhertzmann.com · 03/06/2026

Sometimes I sketch really fast, in ways that produce drawings that seem recognizable, but often with few recognizable low-level features. Often, objects are recognizable but the individaul details don't make much sense. 2/

100

Aaron Hertzmann @aaronhertzmann.com · 26/05/2026

After all, what is the alternative? Maybe you alone are conscious. Or maybe there are a few, undetected philosophical zombies amongst us. Either way, it would seem that consciousness plays no meaningful role in human behavior or abilities, despite shared biology. 4/

100

Aaron Hertzmann @aaronhertzmann.com · 21/05/2026

Now that I've looked up Thorndike's law, I want to be clear that I was not claiming to have discovered it. It is the main axiom of every animal trainer and book I've encountered (all positive-reinfocement based). It probably comes from Karen Pryor, who wrote these relevant words in 1999:

110

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

I’m not sure but it would be some generalization that requires long-term planning in unfamiliar situations. See also the post:

000

Aaron Hertzmann @aaronhertzmann.com · 18/05/2026

The best book on behaviorism for animal training (and also humans) that I’ve read is: Don’t Shoot The Dog by Karen Pryor, the trainer that led the way in positive reinforcement training. I didn't learn how to train dogs, but I learned to better understand the training skills I had learned. 10/

Book cover: _Don't Shoot the Dog_ by Karen Pryor

110

Aaron Hertzmann @aaronhertzmann.com · 18/05/2026

Banishing "knowledge" from your dog vocabulary lets you describe dog behavior better. A dog does not know "he gets a treat if he sits", he has learned the habit of sitting in response to sitting, and to expect a treat. A dog doesn't "know" their name, they have learned to respond to their name. 3/

100

Aaron Hertzmann @aaronhertzmann.com · 18/05/2026

Dogs can be described well behaviorally: they learn from conditioning (including training); they learn associations (this thing good, that place safe, etc.); they have internal states, and so on. Dogs do not know about the future or the past and they cannot plan or solve multistage problems. 2/

Shelter dog Xochi sits while being commanded to sit, with a big excited smile on her face.

100

Aaron Hertzmann @aaronhertzmann.com · 28/04/2026

@rodneyabrooks.bsky.social called them "suitcase words." www.technologyreview.com/2017/10/06/2...

Excerpt from Brooks article:
Marvin Minsky called words that carry a variety of meanings “suitcase words.” “Learning” is a powerful suitcase word; it can refer to so many different types of experience. Learning to use chopsticks is a very different experience from learning the tune of a new song. And learning to write code is a very different experience from learning your way around a city.

When people hear that machine learning is making great strides in some new domain, they tend to use as a mental model the way in which a person would learn that new domain. However, machine learning is very brittle, and it requires lots of preparation by human researchers or engineers, special-purpose coding, special-purpose sets of training data, and a custom learning structure for each new problem domain. Today’s machine learning is not at all the sponge-like learning that humans engage in, making rapid progress in a new domain without having to be surgically altered or purpose-built.

Likewise, when people hear that a computer can beat the world chess champion (in 1997) or one of the world’s best Go players (in 2016), they tend to think that it is “playing” the game just as a human would. Of course, in reality those programs had no idea what a game actually was, or even that they were playing. They were also much less adaptable. When humans play a game, a small change in rules does not throw them off. Not so for AlphaGo or Deep Blue.

Suitcase words mislead people about how well machines are doing at tasks that people can do. That is partly because AI researchers—and, worse, their institutional press offices—are eager to claim progress in an instance of a suitcase concept. The important phrase here is “an instance.” That detail soon gets lost. Headlines trumpet the suitcase word, and warp the general understanding of where AI is and how close it is to accomplishing more.

020

Aaron Hertzmann @aaronhertzmann.com · 30/03/2026

I got 42.64, but I have a lot of practice using HSV color sliders to match colors (from practice with digital drawing), and I can think of a lot of words to describe color that help me remember a color

020

Aaron Hertzmann @aaronhertzmann.com · 25/03/2026

Artists also use conceptual knowledge. For example, a person's eyes are halfway between the top & bottom of their head. But it's so easy to get this wrong, like in this sketch I made! Here's are illustrations of one technique for avoiding the problem, including sketches by Veronese, 1568. 13/

Quick cartoon I drew of a person. Eyes are too high up in the head for a realistic drawing. For a cartoon this is great, but not if this were an initial sketch for a realistic drawing or painting.

How to plan faces as ovals, by William Goeree, 1688

Detail of planning sketch by Paolo Veronese, 1568

120

Aaron Hertzmann @aaronhertzmann.com · 25/03/2026

Learning to draw also involves learning eye-hand coordination skills, like the "Target Locking" behavior shown in this clip from John Tchalenko's film "Capturing Life." This is a common drawing behavior, but people don't know about it because people don't know what their eyes are doing. 12/

160

Aaron Hertzmann @aaronhertzmann.com · 25/03/2026

What happens if you restrict eye movements? Here’s a 30-second "blind drawing" I made, without looking once at the drawing. I took the photo a moment later. The “errors” are typical of blind drawing: individual shapes are accurate, but their proportions and placements are not. 10/

130

Aaron Hertzmann @aaronhertzmann.com · 25/03/2026

These visual limitations explain why artists typically move their eyes so frequently. Here’s a person copying a picture, recorded with an eye tracking device. The black rectangle show where the his eyes are looking at any moment. 8/ (Clip from "Capturing Life" by John Tchalenko)

150

Aaron Hertzmann @aaronhertzmann.com · 25/03/2026

One clue is that human vision is severely limited, both in terms of what we see at any moment, and how much of it we remember. And, we are *totally unaware* of how much we do not see, until it is pointed out. Suppose you wanted to draw a picture from this photo of a houseplant, 5/

A houseplant with several branches and several leaves per branch

110

Aaron Hertzmann @aaronhertzmann.com · 25/03/2026

Here's Henri Matisse drawing a portrait of his son. Watch his head and eyes, and how often they look back and forth between his drawing and his son. Why does he look back and forth so frequently—why can't he just draw his son from memory? And where exactly is he looking? 4/

274

Aaron Hertzmann @aaronhertzmann.com · 25/03/2026

One might think that an artist draws a picture by looking at the world, forming a mental picture of it, and then drawing from that picture. Drawing errors are perceptual or motor errors. But no one draws this way, because no one actually keeps a complete picture of what they see in their head. 3/

Flowchart from source copy to a mental representation produced by perception, to a copy of the drawing made by motor control.

130

Aaron Hertzmann @aaronhertzmann.com · 25/03/2026

Drawing is a creative activity, accessible to everyone: all you need is a pen or pencil and paper, and people have been drawing for at least 45,000 years. Psychologists have extensively studied drawing for decades, but no coherent theory for the process of observational drawing has emerged. 2/

The author sketching a church altar on an iPad

140

Aaron Hertzmann @aaronhertzmann.com · 22/12/2025

(But don’t worry, she gets lots of pampering and adventures.)

050

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

Aaron Hertzmann @aaronhertzmann.com · 12/12/2025

Back before text-to-image, people could appreciate the weird, surreal quality of GAN art. It evoked a lot of historical weird and surreal art, like Francis Bacon or Odilon Redon, without the baggage of today's "AI". See: arxiv.org/abs/1910.04639 .

160

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

Aaron Hertzmann @aaronhertzmann.com · 28/11/2025

I visited Bletchley for a statistics conference in 2006, and expert/hobbyist Tony Sale told us how they had rebuilt a Collossus from standard telephone exchange parts. The original Collossus was destroyed. I think the same goes for the Bombes. Link: www.codesandciphers.org.uk

120

Aaron Hertzmann @aaronhertzmann.com · 24/11/2025

And Battle of the Network Stars!

010

Aaron Hertzmann @aaronhertzmann.com · 24/10/2025

Arguing over whether something is or isn't art is usually a huge waste of time. Just call it art. Instead, discuss its quality, functioning, and values. Is it expressive? Skillful? Ethical? Meaningful? etc. 10/

172

Aaron Hertzmann @aaronhertzmann.com · 24/10/2025

I argue that art is a social behavior, a product of our evolutionary history. We care about art made by people (even when they used computers to do so). Computers are not people, and so we don't accept computers artists. "Computer-generated art" and "AI art" are ultimately art made by people. 8/

130

Aaron Hertzmann @aaronhertzmann.com · 24/10/2025

Plus seminal conceptual artworks like Sol LeWitt's "Wall Paintings" and John Baldessari's "Commissioned Paintings", where the artist didn't make the painting. 4/

A 1969 artwork by John Baldessari, one of his "Commissioned Paintings," painted by a hired worker

111

Aaron Hertzmann @aaronhertzmann.com · 24/10/2025

A common misconception is that "a computer that makes high-quality pictures is an artist". But we have 60 years of experience of computer algorithms and tools that can make high-quality images. 3/

1102

Aaron Hertzmann @aaronhertzmann.com · 15/10/2025

For me, relating my drawings to scientific theories of drawing happens after drawing, not during ("what did I do?"), but it has really enriched my drawing practice and also inspired nearly all of my published research in the past 5 years. psyarxiv.com/pq8nb

061

Aaron Hertzmann @aaronhertzmann.com · 07/10/2025

He does seem to agree that GenAI is the latest stage in this culture.ghost.io/genai-is-our...

010

Aaron Hertzmann @aaronhertzmann.com · 07/10/2025

"People don't make good pop music anymore like they did in the 1950s, it went bad after that" aaronhertzmann.com/2022/12/16/s...

100

Aaron Hertzmann @aaronhertzmann.com · 07/10/2025

In _Status and Culture_ (chapter 10), W David Marx summarizes claims of cultural stasis from the 2010s due to Internet culture. But I think, whenever one thinks "nothing has moved me lately," the first question to ask is "is it me?" It gets harder to appreciate new culture as you get older.

131

Aaron Hertzmann @aaronhertzmann.com · 01/10/2025

In fact, Sousa's own band made a lot of money on recordings, and he himself had profited from reusing Gilbert and Sullivan's work without compensation. He wildly overstated the case claiming that recorded music has no soul, but also correctly predicted some things we've lost. end/

130

Aaron Hertzmann @aaronhertzmann.com · 01/10/2025

But he also correctly predicted a decline of social music-making. People used to make music together as a social activity much more than now, when so much of our entertainment is passive. But social music-making is probably the reason we have music at all. 3/

151

Aaron Hertzmann @aaronhertzmann.com · 30/09/2025

More research is needed to understand size perception, but it seems to be partly contextual, partly related to visual angle (i.e., no focal length compensation). jov.arvojournals.org/article.aspx... scholar.archive.org/work/eyreite...

020

Aaron Hertzmann @aaronhertzmann.com · 24/09/2025

This seems similar to two illusions by @akiyoshikitaoka.bsky.social , but simpler (which is good). aaronhertzmann.com/2024/05/24/f...

031