Brian Eno with a thoughtful essay (a talk, actually) on how he got interested in generative systems: http://www.inmotionmagazine.com/eno1.html pic.twitter.com/VoVKnbNNxR
Show this thread
-
I was watching the documentary sound breaking this weekend and it gave me all sorts of ideas about generative models in ML. I'm surprised the analogy of a "Producer" hasn't been explored more (with (conditional) GANs being instruments in analogy)
@ahandvanish @RebeccaFiebrink @lukexi or @shancarter might have done some such work, or know of it?
-
-
Perhaps related to the interactive GANs work by Alexei Efros's group. And to this piece
@shancarter and I wrote:https://distill.pub/2017/aia/Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I think it’s implicit in work but I haven’t seen it explicitly framed as such. Would “suggested edits” count in your book? Like advanced inline spell check?
-
Not sure what
@KyleCranmer had in mind, exactly. That reminds me of@robinsloan's beautiful experiments (I don't immediately recall the link - wasn't there a video at IO or something similar?) - 3 more replies
New conversation -
-
-
@ada_rob do you know of any generative models that serve the role of a music “producer”? -
We've discussed that quite a bit in Magenta, but I don't believe anyone has done it yet. A primary difficulty would be getting training data.
- 7 more replies
New conversation -
-
-
I haven't, but that's an interesting + intuitive idea!
- 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.