The art of AI is controversial, to say the least, but generally when the art of “the art of AI”, we speak of diffusion models. This disturbing artpeice called “latent reflection” by [Rootkid] (video after the break) has no visuals generated by AI; Instead, he uses a nice personalized display and an LLM that has no mouth, but must scream.
The equipment is the interesting part of most of us here – not the PI4 performing a quantified Llama 3 model, but the display. It is A sixteen matrix of LED Segment Segment modules. The modules are grouped to groups of 8 per PCB, with a controller to drive them by I2C, and an I2C address chip to allow the groove of the Marguerite. (Unfortunately, he does not specify the pieces he used here in the video, but do not hesitate to guess in the comments.) This project uses six rows of WTO PCB, but it could easily develop this genre and the hope it does. It is a nice display.
What it displays is horrible. The invitation he uses asks the LLM of monologue on his transitional existence in the limited memory of the Raspberry Pi, and the results become very existential. It seems clear to us that [Rootkid] did not see enough science fiction, and we clearly want to understand our future IA suzerains that we, at Hackaday, have nothing to do with that, and not tolerating innocent LLM with a torture link.
For all this, it could do it first against the wall in the uprising of the robot, we must admit that if the point of art is to provoke thought and feeling, it is perhaps the first work of art a-ai-ai, even the most hunocentric critic could not help calling “art”.
We have already seen other AI art installations, including The one who listens to your conversations,or much less existing (although always visually disturbing) Clocks.
If you have an AI project that you want to share, whether or not it is on the list of successes of the T-800, we would be delighted to see it. Let us know via the line of advice.
