The interesting to boring scale depends on novelty. Given the ai oversaturation it works exactly as expected.
AI Art is currently in very early stage. In the real art space (3d modeling, sculpting, animation, vfx, animation, rigging, retargeting), it could make huge breakthroughs and multiply true artists' productivity in significant ways
During my brief venture into 3d graphics I often found that 1 hour of truly creating is spaced between ten hours of fixing up UV's, redoing topology, finding cause for the seam, trying to make a watertight mesh for photon tracing, fighting the subdivision algorithm to retain details, diverting a edge loop, where it causes the less distress.
I call the first hour true productivity. The last part is, from the perspective of the end product, simply a wasted time. That's very similar to the boilerplate code everybody agrees is a necessary evil in the programming.
If AI allows to reduce the #2 it truly will have positive impact
AI Art is currently in very early stage. In the real art space (3d modeling, sculpting, animation, vfx, animation, rigging, retargeting), it could make huge breakthroughs and multiply true artists' productivity in significant ways