I heard a good quote a few weeks ago about AI, because LLMs need to learn, what they learn comes from historical models, which are then used to extrapolate behaviour moving forward with what they have learned.
Case in point, to draw, AI had to look at pictures of drawings to define structure and learning in order to draw in a similar model.
I pick up a pencil, stare at the piece of paper, and see what comes out. Now, I would love to have the benefit of a model of all these artists in my head to help me confirm the direction and path I want to draw, but I don’t. So I scribble on the page and see what comes out, maybe it’s a new style, maybe it’s not, maybe it’s some rehashed combination of someone else, just like AI.
But every now and again, you get that drop of something new, maybe someone else hasn’t done it yet, maybe they have, and they thought it was nothing, but it was that brand new spark of an idea or concept that you tried out.
I don’t think AI has found that spark yet. We’re pushing it there as fast as we can, because, well, I don’t know what will happen then, but it’s not there yet.
It hasn’t found the direction to move forward.
We, on the other hand, still have the ability to move forward and always will, and sometimes that’s the comfort of when toiling away at the keyboard.