Sora, in imitating art, also subtly discourages it. How? Well, in ordinary movie making, various artists collaborate to channel a vision into video. The instrumentation (and sets, etc.) they use to do this entails millions of degrees of freedom. They have freedom to create that goes beyond any words that may describe it. Pre-trained generative models have many fewer degrees of freedom. For instance, try asking Bing t…
If I have a vision, I cannot pull that out of my head and make it real using generative AI. Maybe I can get someone else's take on my concept, but I don't have the flexibility to build the image in my head. That's the real difference between the new tools of old and the "new tools" of today. You can't actually create with them, just sift and hope with varying degrees of control, but no creativity.