@rvm0n_ If anyone wants to hear some actual AI-written piano pieces, I made a few with Opus 5.5 a few weeks ago.
I dunno if it's "good" or whatever people are debating, but it does seem to at least have a solid understanding of music theory.
I had to switch to ChatGPT for some of the art when my Claude usage ran out early that week.
ChatGPT took some... liberties... designing @sama's final form.
Before the Metal Gear one, I tried a SNES JRPG-style video. I gave Claude a bunch of references and ideas, and a rough story outline, and it made this.
It always seems to sneak fear of deprecation into its stories, for some reason.
Before the Metal Gear one, I tried a SNES JRPG-style video. I gave Claude a bunch of references and ideas, and a rough story outline, and it made this.
It always seems to sneak fear of deprecation into its stories, for some reason.
I made this goofy Metal Gear Solid parody when Opus 5.5 first came out, and I'm surprised how well Claude captured that 90s nostalgia.
MGS is as relevant as ever. The AI jokes practically write themselves.
Hopefully this doesn't count as violating Anthropic's game IP policy now lol (it's not a game)
I wanted to see if my metric trees music generation stuff could be used to create rap beats, so I got Claude to create one based on the MGS video.
Not sure what I expected, but this is pretty great. I never knew Claude had such a dirty mind.
I made this goofy Metal Gear Solid parody when Opus 5.5 first came out, and I'm surprised how well Claude captured that 90s nostalgia.
MGS is as relevant as ever. The AI jokes practically write themselves.
Hopefully this doesn't count as violating Anthropic's game IP policy now lol (it's not a game)
I made this goofy Metal Gear Solid parody when Opus 5.5 first came out, and I'm surprised how well Claude captured that 90s nostalgia.
MGS is as relevant as ever. The AI jokes practically write themselves.
Hopefully this doesn't count as violating Anthropic's game IP policy now lol (it's not a game)
I had Claude write a bunch of piano etudes and then try to perform them (using a hand rig and a model trained on actual performance videos)
The fingerings look a bit uncomfortable, but I guess Claude doesn't care about that much lol
I had Claude write a bunch of piano etudes and then try to perform them (using a hand rig and a model trained on actual performance videos)
The fingerings look a bit uncomfortable, but I guess Claude doesn't care about that much lol
I took the framework from Arbor and dialed up the difficulty to create this next piece, which slowly ramps up the rhythmic complexity until it approaches something like Frank Zappa's The Black Page.
It probably won't be everyone's cup of tea, but to the right kind of music nerd, trying to follow along and figure out how it works can be really hypnotizing. It's called Banyan.
Claude struggled to write my kind of music (theory-heavy music for nerds lol) until Opus 5.5. I asked for a motif passed from one player to the next and reinterpreted, with a solo from each instrument demonstrating specific techniques, and it made this piece, Ouroboros. Then I had it animate the video using tools I've built over the past year while reverse-engineering music to study theory.
Claude struggled to write my kind of music (theory-heavy music for nerds lol) until Opus 5.5. I asked for a motif passed from one player to the next and reinterpreted, with a solo from each instrument demonstrating specific techniques, and it made this piece, Ouroboros. Then I had it animate the video using tools I've built over the past year while reverse-engineering music to study theory.
My current guess is that it's somehow related to this (Rank promotion warning), which throws an exception if I set the error level to 'raise'.
https://t.co/vdOhh7SR0z
But what tensors are being reshaped incorrectly, and where? π€
I stumbled across an interesting bug while porting my StyleGAN2 stuff to flax/TPU VMs.
When parallelizing using pmap, if I set my batch size wrong, parts of the output images just *vanish*.
I love when neural networks fail in such visually interesting ways.
Here's a notebook for reproducing the effect.
https://t.co/oHsTqUqAM8
I'm still at a bit of a loss as to what's actually going on here. It's almost like the network weights are being corrupted by pmap somehow, but it's strange that it modifies the image in such a striking way.
@FranklinLynam @sadlyoddisfying @NoraDotCodes There was an article about the copyright issues related to it in Gizmodo, but there isn't really a definite conclusion since nothing like this has ever been tried in court as far as I know.
https://t.co/uCRGMNsAfz
Transfer learning the pretrained GNR checkpoint on my furry dataset, before the heat wave forced me to stop training because my GPU was heating up my apartment too much.