Meandering Tour

Painting, Remote Taste Test, Artificial Intelligence, Murmuration, and Shrimp Welfare

5) Late onset artistry

When I was in New York a few weeks ago, I went to see 300 Paintings, a one-man show by Sam Kissajukian, a former stand-up comic who, in his 30’s, out of the blue, and in a manic state, began painting large canvasses in an abandoned factory. On stage, Kissajukian amusingly narrates a slide show of his resulting works while exploring the relationship between his art, creativity, and mental health.

After the show, I spoke with him in the lobby, where he was exhibiting some of his paintings. I asked if he could explain such a sudden artistic breakthrough, with no experience. His answer was facile, but not disingenuous, as far as I could tell. He said he knew how to express himself through stand-up comedy and just redirected his imagination into painting.

I can’t fathom the rapid mastery of such a creative endeavor — the brushwork, design, coloration, and hand-eye coordination required to render a bowl of fruit much less a human face. For much of my career, I wrote strategic planning PowerPoint decks (full of graphs and charts featuring very pleasing pastel colors, if I must say so myself). I once tried to paint a watercolor seascape, but it wound up looking like something out of a third grade art class.

I can, however, spell and punctuate reasonably well, so I’m trying my hand at this Substack.

4) May I have a byte, please?

The Internet has revolutionized our ability to share images, video, audio, and even a sense of touch remotely. Now comes e-Taste, a device developed by scientists at Ohio State University, which enables the wireless transmission of taste sensations.

Writing in the journal Science Advances, the researchers left me hungry for an explanation I could actually understand, so I turned to a newspaper, which explained the two-part device as follows:

The first is a sensing platform – or “electronic tongue” – that captures data about the concentration of five different taste chemicals in a liquidised sample of a food or drink and sends the information wirelessly to the second component. This device uses small electromagnetic pumps to push liquids through a layer of gel containing five edible taste chemicals, allowing their delivery to an outlet, placed in a recipient’s mouth, to be controlled.

The five taste chemicals, which can be mixed before delivery, correspond to those detected in the original consumable item: glucose, which creates a sweet sensation; citric acid, which produces a sour taste; sodium chloride, which generates a salty flavour; magnesium chloride, a bitter taste; and glutamate, which elicits the savoury taste known as umami.

The team adds that the system also allows flavours to be delivered to specific regions of the tongue – something they say could be used to investigate human taste perception.

What? Flipping back to the actual journal article, titled “A sensor-actuator–coupled gustatory interface chemically connecting virtual and real environments for remote tasting,” I was thrilled to discover the following illustrations that I hoped might clarify how this could be possible. Nope.

I next turned to Science Daily, billed as “Your source for the latest research news,” but was stopped short by this sensational headline, which struck me as overselling at this point:

New device could allow you to taste a cake in virtual reality

From fish soup to coffee, 'e-Taste' delivered, study finds

The Buckeyes will not be deterred. "The chemical dimension in the current VR [virtual reality] and AR [augmented reality] realm is relatively underrepresented, especially when we talk about olfaction and gustation," said Jinghua Li, co-author of the study. "It's a gap that needs to be filled, and we've developed that with this next-generation system."

Dippin' Dot-Coms, anyone? Microchips? Silicon wafers? Cookies? I’ll stop now, as my old friend Glenn would say.

3) Machine learning

Many of us have tested artificial intelligence programs (aka Large Language Models, or LLMs) such as OpenAI’s ChatGPT by asking them to compose poems, write short stories, or answer trivia questions using simple prompts. On a more sophisticated level, I’ve read about AI receiving high marks from professors on graduate school exams and taking over coding assignments from teams of programmers.

But it’s been difficult, for me at least, to appreciate the true power of machine learning until I read about a recent case.

Stephanie Kelton

, who writes The Lens on Substack, cited a simulation run by the Bank of Canada to assess the potential impact on the country’s GDP of a 25% tariff on all Canadian exports to the United States. Its answer: a decline of 6%.

She wondered if Deepseek, the new, much less expensive Chinese competitor to ChatGPT, would reach the same conclusion, so she asked the following question, including a request for an actual model and simulation:

Quickly scan what Deepseek achieved. You don’t actually have to read it; just scroll through the maze of numbers, equations, and calculations. Deepseek’s answer was roughly in line with the Bank of Canada’s projection.

“I wonder how long it took the staff at the BoC to come up with their figures,” asks Kelton. “It took Deepseek about 12 seconds.”

I realize some of the more informed readers here (I’d be interested in your comments) will point out that this is a relatively simple problem compared with the massive complexities AI already is tackling. But for the uninitiated, I think this is a good example of the enormous potential of LLMs, both positive and negative.

Hey, wait a minute! Let’s ask Deepseek the following question, “Can you transmit tastes over the Internet.” Answer:

Currently, it is not possible to transmit tastes over the Internet in the same way we transmit visual or auditory information. Taste is a complex sensory experience that involves not only the taste buds on the tongue but also the sense of smell, texture, and even temperature. These factors make it challenging to replicate and transmit taste digitally. However, there have been some experimental efforts to simulate taste experiences [e.g.,]

1. Digital Taste Interfaces: Researchers have developed devices that can simulate basic tastes (sweet, salty, sour, bitter, and umami) by using electrical and thermal stimulation on the tongue. These devices can be controlled digitally, but they are still in the experimental stage and not widely available.

. . . .

Despite these advancements, transmitting taste over the Internet in a way that is as seamless and immediate as transmitting images or sound remains a significant technological challenge.

Does Item 1 above suggest that Deepseek has read about the Ohio State research or missed it? I followed up with, “What about the new research at Ohio State University?”

So it does know about it! But the rest of Deepseek’s answer, right or wrong, does not credit the OSU team with the kind of specific success, e.g. fish soup and coffee, reported by Ohio State News.

From time to time, I do miss old-fashioned research in real stacks.

Source: Library of Congress

2) Sidetracks

• Seinfeld anticipated everything, even Ukraine and Greenland. This scene is from Season 6, Episode 12 in January 1995.

• Murmuration

• Ten years ago this week, the dress color controversy hit the Internet.

• Tuesday night’s presidential speech to Congress was not technically a State of the Union address. (h/t my Meandering Muse, Suzie)

• The winners of the 2025 World Nature Photography Awards were just announced.

• A reader in Palm Beach, FL spotted this license plate last week:

1) A real pain

I must spend half my life at home wondering what my dogs Adley and Luna are “thinking.” Do they experience feelings like joy, sadness, and fear? Do they love me? I don’t care what the research says — of course they do!

But many assume that so-called “lesser” species are not sentient. Not so fast.

A recent New Yorker article explored whether insects can feel pain. Researchers apparently found “strong evidence for pain” in adult flies, mosquitoes, cockroaches, and termites, using “motivational trade-off” tests, among other techniques. In a surprising sidenote (I might even call it “meandering” as New Yorker features tend to do), I learned that in 2022, the UK passed the Animal Welfare (Sentience) Act that recognizes animal sentience in law for the first time, including all vertebrates and some invertebrates such as octopuses and lobsters.

And then a few weeks ago, NPR reported on lab mice giving “first aid” to unconscious mates. "To me, this looks very much like a behavior that's driven by what I would call the altruistic impulse," says a neuroscientist who was not involved in the study. "We can't infer just from our observations that these mice have an intention to help. We only know that they're responding to an animal in need and they perform a behavior that does benefit them."

What about shrimp? Glad you asked. According to the January 2025 issue of the quarterly online journal Asterisk, there is a growing body of evidence that they do experience pain. The author, Andrés Jiménez Zorrilla, who discloses he is Chairman of the Board and co-founder of the Shrimp Welfare Project, writes:

The Welfare Range Table, a comprehensive assessment of animal capacities, reveals that shrimp display numerous sophisticated behaviors: they can learn from experience and show anxiety-like states when threatened; when injured, they display targeted grooming behavior (basically, tending to the injury) that can be reduced by local anaesthetic, suggesting conscious pain response rather than mere reflex; and their nervous systems show a striking level of complexity, with clear evidence of integrative processing centers that could support conscious experience.

Zorrilla calls for better treatment of shrimp in their supply chains, including an estimated 230 billion shrimp in aquafarms at any given time, much like Temple Grandin advocated for the humane treatment of livestock headed to slaughter.

I don’t know how to fully process such studies, but all of this is food for thought thought-provoking.

Why don’t we just finish up by coming full circle — to a video of a dog painting, with no formal training.

Bonus Track

Wentzle Ruml IV was an original member of the legendary L.A.-based Zephyr skateboard team known as the Dogtown Z-Boys. With their surf-like style, they revolutionized skateboarding in the mid-1970s; their aerial moves, often performed in empty swimming pools, ushered in the era of extreme sports. This trailer, from a 2001 documentary, will give you a sense of who they were and what they accomplished. I met Wentzle in the summer of 2020 when he was shingling a cottage on the edge of a high ocean dune on Cape Cod, MA. He made an impression and inspired this song. My friend Rick says it sounds more like a sea shanty than a fitting speed metal track from the 1970s. That’s true, but I met Wentzle 45 years later on a quiet New England beach.

Wentzle and me