JOURNALING PROJECT No. 1 – IMAGE & INTENTION: A MESSAGE WITHOUT A CODE

When considering the different configurations for the plating of a dish or when imagining the aesthetics of seven or more different ingredients as a finished product, sketches and the imagination can really only go so far. For my project I am experiemnting with AI image generation software because it sits in a strange pocket of the uncanny valley especially in terms of food. The ability to generate a seemingly photo-realistic imagination of the description of your course is helpful in one way, but ChatGPT or ClaudAI will never be able to taste or see an ingredient at all, which bleeds through into the image generation heavily. Lets take a close look at this example AI image which I fed the doccument I have been brainstorming my 5 course meal ideas into.

Certain things will catch your eye immediately: To start, the labels look absolutely insane and list odd mixed letters for the top two courses, labeling them both as the first-bite. As we know, text is still and will likely forever be, the achilles heel of AI tools. It's really unsettling, like seeing a chamelion try to blend in on a checkered tile floor, something so certain like text is so hard to mess up that it triggers that uncanny feeling when confronted with AI. It attempts its best guess at what english words might look like, but it is confused since its been trained off of so many different fonts and graphics, it almost tries to adverage out between every word its ever seen. Barthes notes the importance of the linguistic message through human civilization retrospectively. When we look to the future for text from AI and consider Barthes' message, it frames the importance of text even in a world rushing as fast as it can towards a sci-fi version of our future. "Today, at the level of mass communications, it appears that the linguistic message is indeed present in every image: as title, caption, accompanying press article, film dialogue, comic strip balloon. Which shows that it is not very accurate to talk of a civilization of the image - we are still, and more than ever, a civilization of writing, writing and speech continuing to be the full terms of the informational structure." (Barthes, p.38) Even AI jumps to title things and categorize them with explicit text since its been trained off of decades of online information generated by humans who did the same. This undermines anchorage and introduces dissonance. The text isn't the only strange thing, the denoted image is in itself, completely off base. My five course meal has been interpreted by the AI to be 4 courses, putting the desert course (Halved White Grape–Preserved Peach in Chamomile-Honey Reduction) on the same plate, nice and cuddled up next to the main course. The denoted image is technically where AI succeeds. It can certainly mimick the look of culinary photography, but it certainly lacks a human intentionality, it's a hollow success. The texture of the rice is certainly strange and almost hallucinated by the machine to look porridge like, and while its 94% realistic or so, the imperfections bleed through. The connotated image is the tricky part, the AI (without any particular input to lead it to this conclusion, through context alone) attempts to replicate the culture of fine dining and the gormet. With the specific plating, portion size, and ceramic choice, it tries to invoke a high end feel conveying the message through the ratio of the food to plate as simple as that seems. Since this AI picture exists nowhere in space and requires no camera or further camera man, we can learn alot about the connotated image by what lighting and angle the AI chooses too. It tries to immitate the editorial photography of a micheline-starred restaurant and the composition reads the same too. If there is one thing I've learned throuhgout the rapid application of AI to so many facets of life which it has no buisness being in, it is that AI should never be used to generate a final product, that is, it is amazing at getting the ball rolling and bouncing ideas off of, but it will never make something human and putting more AI slop onto the internet is just fuel for the flame. This is why I have been using it to act as a heightened version of my imagionation for how these dishes could eventually look. As for a link to my checkpoint 1 slideshow here you go:https://docs.google.com/presentation/d/1WIcMBW9jJ8aTHjYnGuK95p1-xPou_0Mh4HatVke4T7g/edit?usp=sharing