There was a Reddit thread (I know, I know) estimating the amount of compute potentially required for Sora, and the guess they arrived at was it might need its own nuclear power plant lol.
If you listen to a few of the higher-up AI people like Sam Altman, they seem to all suggest the next great battle will be for energy. It's likely our tools will quickly outgrow the available power to run them.
I feel like Alberta is ripe for nuclear. No fault lines, no tsunamis... I say we go full Springfield.
The deepfakes are getting incredibly good... although I feel the Argil.ai ones are leveraging existing videos and using AI for the voice match and lip syncing. Lip syncing is getting insane. Regardless of how it's done, this could probably get out of hand real quick (although it'll be great for movie dubbing).
Udio (Suno competitor) has launched, and I feel like the recording industry is going to have puppies. They literally use lyrics from House of the Rising Sun in their marketing. The voices sound incredibly good so far, but I've only dug in a little. Accounts are free for the beta period, and you can make 1200 songs/month. I'm currently trying to create something, but I have a feeling the site is getting demolished by user traffic.
Udio is potentially more legally problematic that Suno in that apparently you can add artists to your prompt and get that style. It feels as though we're not far off from being able to take that band you love that went in that weird direction (that you didn't love), and create the album you never got. That's... I don't know what I think about that.
The AI is positively bonkers right now. Products like Suno launch and then have "Suno-killers" launching 3 weeks later.
Udio (Suno competitor) has launched, and I feel like the recording industry is going to have puppies. They literally use lyrics from House of the Rising Sun in their marketing. The voices sound incredibly good so far, but I've only dug in a little. Accounts are free for the beta period, and you can make 1200 songs/month. I'm currently trying to create something, but I have a feeling the site is getting demolished by user traffic.
Udio is potentially more legally problematic that Suno in that apparently you can add artists to your prompt and get that style. It feels as though we're not far off from being able to take that band you love that went in that weird direction (that you didn't love), and create the album you never got. That's... I don't know what I think about that.
The AI is positively bonkers right now. Products like Suno launch and then have "Suno-killers" launching 3 weeks later.
Goodbye streaming services. Right now you can upload songs to all the services with no specific quality validation. Imagine bots creating and uploading thousands of songs a day then spamming social media in the hope that one catches on and makes a ton of money. I imagine the streaming services will need to crack down on this somehow.
Goodbye streaming services. Right now you can upload songs to all the services with no specific quality validation. Imagine bots creating and uploading thousands of songs a day then spamming social media in the hope that one catches on and makes a ton of money. I imagine the streaming services will need to crack down on this somehow.
At some point I think we simply drown in all the content. Or the idea that there's a general pool of content we all pull from becomes a thing of the past. The only pool we draw from is ours, and every book/song/show we consume is generated for us specifically.
I love music, but throughout my day I more of less listen to the same 7 hour long movie score orchestral soundtrack playlist over and over again. I'm rarely even aware it's going, it just locks my mind into a focused groove.
I was playing around with making that type of music in Udio yesterday, and it's to the point I could probably listen to it all day long and it would likely do what I need. If we could get it to a place where it could create a 2 minute loop in under 2 minutes, it could conceivably run all day and I'd barely notice it was always a few seconds ahead of me.
This isn't an ideal direction for music mind you, but it's a use case that would serve me specifically.
Found this interesting, book written with the help of AI but edited by a person granted a kind of copyright:
"The USCO’s notice granting Shupe copyright registration of her book does not recognize her as author of the whole text as is conventional for written works. Instead she is considered the author of the “selection, coordination, and arrangement of text generated by artificial intelligence.” This means no one can copy the book without permission, but the actual sentences and paragraphs themselves are not copyrighted and could theoretically be rearranged and republished as a different book."
Found this interesting, book written with the help of AI but edited by a person granted a kind of copyright:
"The USCO’s notice granting Shupe copyright registration of her book does not recognize her as author of the whole text as is conventional for written works. Instead she is considered the author of the “selection, coordination, and arrangement of text generated by artificial intelligence.” This means no one can copy the book without permission, but the actual sentences and paragraphs themselves are not copyrighted and could theoretically be rearranged and republished as a different book."
Is there a free AI tool that can sing text to song? We had chat GPT make up a song for a departing co-worker, but no one brave enough to try to sing it
