Interior With Sudden Joy

by me

“Early Snow – Rhinecliff Hotel”, 2017, oil on canvas

“Durham, August 14, 2017”, 2017, oil on canvas

American painter Celeste Dupuy-Spencer passed away last week at the age of 46. The images above are from her 2017 exhibition Wild and Blue at Marlborough gallery in NYC. She had also been part of the Whitney Biennial earlier that same year.

The first painting is of the Rhinecliff Hotel, a bar she frequented while growing up in Rhinebeck, NY. The second, Durham, August 14, 2017, is of the metal confederate statue that protesters tore down that year. That painting was also included in her section of Made in L.A. 2018, Hammer Museum‘s biennial exhibition of artists from the Los Angeles area. Her later work was often very political, including several paintings that are dense with imagery.

In this 2018 Bomb magazine interview, Dupuy-Spencer discusses some of her past struggles and provides insights into her practice.

The sea ponies appear to be starting a summoning.

Or maybe it’s just a conga line.

Julie Gough (b1965) Trawlwoolway people/ Tasmania, Australia

The Promise II (2018-21) chair, kangaroo vellum, LED

A.I. chatbots provide a convincing illusion of consciousness, but we know they are just a sequence of lifeless math calculations. They are no more conscious than the horse in Muybridge’s animation is alive. The main difference is that people could easily look inside Muybridge’s machine to see the photographs underlying his movie. Most people don’t look inside A.I. chatbots. But if they did, they would see that it’s essentially the same story—an A.I. chatbot is just a mathematical flip-book. All A.I. systems today are built around neural networks, which, despite their fanciful name, are just math formulas. That’s right: At the heart of Claude, ChatGPT, and the rest is a math formula like f(x) = ax2 + bx + c, except much, much bigger. Instead of having just three parameters—a, b, and c in this quadratic—the latest A.I. chatbots are estimated to have trillions of parameters. But the size doesn’t change the fact that they are mere formulas: You plug in a value for x, and out comes the value f(x). I think it’s safe to say that a math formula written on a sheet of paper is not a conscious entity—just as a croissant recipe is not an edible pastry. Perhaps plugging numbers into this formula brings it to life, just as the baker brings a recipe to life by mixing ingredients in a bowl and baking them in the oven? No, the analogy can’t take us that far; there are no ingredients here, just a calculation you could, in principle, do by hand. If you think writing out a math calculation with pen and paper is enough to create consciousness, you have far more faith in mathematics than I do. And writing the formula in computer code doesn’t breathe life into it; it just uses a different set of symbols to describe it. There is no consciousness to discover here when you break down what is inside the machine that is A.I.—just formulas. When a chatbot isn’t actively responding to your prompt, it sits around waiting, doing absolutely nothing. Not thinking, feeling, experiencing self-awareness or consciousness or anything else. Until your prompt comes in, the chatbot is just a big fancy recipe, vacantly waiting for someone to do something with it. When your prompt does arrive, the chatbot merely translates it to a number, pipes this number through a giant formula, then translates the resulting number back to text. That’s all.