Good and bad ideas both come from the same fountain of speculation and experiment.
It's only a very small percentage of creative thinking that ends up connecting with a wider audience, and even then, any success is quite unpredictable.
I always overwrite - really awful, long bits of script - and then I trim it down to the bare bones and then add a little bit to colour it in. At the end of all of my stories, I test for wordless comprehension. So I remove the text and see if it works by itself. And if it does, I feel that that's a successful story.
As a younger person, I was obsessed with Ray Bradbury, and I think his stories did more to shape me as a storyteller than anybody else - even though, when I read them now, a lot of them seem overly sentimental. But that's probably the writer that I've thought about the most, even though I don't necessarily like a lot of his work.
My friend Markus Zusak wrote a story from the point of view of death, 'The Book Thief.' I thought that's a great idea, where your omniscient narrator is death. I'm glad he had that idea because I wouldn't have been able to work so well with it.
The detail adds an element of unexpected something. All fiction is false; what makes it convincing is that it runs alongside the truth. The real world has lots of incidental details, so a painting also has to have that element of imperfection and irregularity, those incidental details.