Definitions of LLMS

This is my research  🛋🔎 Short Version.

LLM = Social Technology 
Diffusion model/multimodel = Aleatory performance art (Speculative Executive)

For the rest of us though, What are LLMS? Some clarifications

Reach artwork has dome some for designers

I'll place this one with the com copywriters/graphic designers to look up. First off within creative industries. Hate it or love it, I've been trying to get to the heart of this nondeterministic praxis 2024.  With the executive dysfunction

"I’d like to propose that we think of generative AI as a technique for producing aleatory artworks. Much of the angst around AI, to me, comes from observing the stochastic, indeterminate, unpredictable nature of it—and framing these things as flaws. But in aleatory art, these qualities are useful and productive: They invite the artist to produce works through different processes, which lead to different aesthetic outcomes."

"The term aleatory refers to anything dependent on random or stochastic processe (like rolling a die, or using the result of a random number generator). Thanks to the avant-garde composer Pierre Boulez, the term was often used in the 20th century to describe art that used chance to create, produce, or perform a work."

Aleatoric performance art is an interesting interpretation of the mechanics.

(You can ignore that Emily Bender paper btw)

So LLMs are non-deterministic?

"LLMs are nondeterministic. Even with temperature 0 they may not respond to the same prompt with the same output — and we deliberately keep temperatures well above 0. Even if they were deterministic, unlike software, there is generally no fixed or limited set of inputs; each prompt is different, anything might go into them, and thence through the black-box neural networks of embeddings & attention heads & skip layers. LLMs are stochastic: their outputs follow unpredictable probability distributions rather than reliable algorithms. They are, again, nondeterministic.

There was a little bait-and-switch above, a rhetorical magician’s trick. Yes, the hardware will run your software in the same way every time ... but the software you first give it is always wrong! When you write code, the outputs are not deterministic in any way that is meaningful to the author. Instead the outputs are truly random… because code has bugs, and you don’t know where those bugs are, or what they do. Software is deterministic in principle, sure. But software in practice is actually far more chaotic than LLMs!"

With this presentation drop box by Henry Ferrel.