You might have seen this somewhere in your timeline. It's now possible to add variables to your prompt which can help tremendously in speeding up your prompting proces and iterating for example.Have you tried it yet?

Hey everyone, I wanted to open up a conversation about the 'look' of our work. In my past life as a photographer and filmmaker, I was obsessed with "realness." That digital, clinical perfection was usually the enemy. We spent our time adding grain, hunting for the right LUTs, and emulating old film stocks in post-processing—anything to give the image some soul. Now that we’re all deep into AI, that "plastic," over-polished look is everywhere. It’s technically impressive, but it’s often a dead giveaway that it was made by a machine. Usually, you’d fix this in post-processing (Lightroom, Photoshop, etc.). But what’s interesting about the node-based tools we’re using now is that we can bake that character in during the generation process. I’m curious to hear from the community: What’s your go-to move for breaking that AI glassiness? I’m currently mapping out a few different workflows to show exactly how you could handle this, so stay tuned for those. But for now how are you guys doing it?