all market commentary for entertainment only help save open-source AI: @thepirateface RT is not endorsement
Trenches
Joined March 2009
- Tweets13.2K
- Following5K
- Followers3.6K
- Likes10.6K
manulcapital retweeted
I’ve been prophesying this for years
Spotify just announced a feature that could either save the music industry or destroy it.
Starting as early as Q1 2027, Spotify Premium subscribers will be able to use AI to transform any participating artist’s song into a completely different version. Turn a folk ballad into a dance track. Make a hip-hop song sound acoustic. Generate a cover where one artist’s voice performs another artist’s song. Remix a pop hit into lo-fi study music. Create the version of your favorite song that only existed in your head until now.
The remixes will be uploaded directly to Spotify. They’ll appear in playlists. They’ll count toward the original song’s chart position. And most importantly the original artist will earn royalties every time someone listens to a fan-created remix of their work.
Spotify has already signed a landmark licensing deal with Universal Music Group the world’s largest record label. Negotiations with Sony Music and Warner Music are ongoing. The feature will cost an additional $4–$5/month on top of Spotify Premium. And analysts project 7–8% of American subscribers will adopt it by 2028.
This is not Suno or Udio generating music from text prompts. This is fans remixing real songs by real artists with the artists’ permission, on the world’s largest streaming platform, with royalties flowing back to the creators.
A music industry analyst who’s covered streaming economics for 15 years told her team something after reading the licensing agreement that reframed how she understood the future of recorded music.
“Spotify just turned 640 million listeners into 640 million potential remixers. The question isn’t whether people will use it they will. The question is whether paying artists for fan remixes solves the economics of streaming or creates a world where the original recording is worth less than the infinite remixes generated from it.”
Here’s what Spotify’s AI remix feature actually is, who wins, who loses, and why this might be the most consequential decision in the music industry since the launch of streaming itself 🧵
manulcapital retweeted
Yep, called it early.
Everyone is trying to solve memory by over engineering, but I think we are gonna get bitter lessoned here too.
All we need is to give agents ability to operate on file system + provision SQL databases.
The main system prompt tells the agent that it can create files for unstructured memory and SQL databases for structured memory. It can index it all in memory index with instructions on what to remember.
Memory index gets injected into system instructions, and it can only be 5k tokens max. Throw an tool call error + hint to reorganize memory index when it goes over 5k tokens.
Agent can now mange its own memory based on what user wants it to remember.
I think models today are good enough to do this now. I am going to try this and report back.
@worklayer already lets agents access both files+databases. If this works, I will get getting rid of the custom memory extension we had created.
manulcapital retweeted
Pfizer lost over a billion dollars on their inhaled insulin product, Exubera. There were a bunch of problems with it (most importantly that it cost twice as much as other forms of insulin) but the funniest was that patients really didn't like using it in public.
You popped a cartridge into it, the clear chamber filled with white "smoke", and then you cleared it one breath. It looked exactly like you were smoking weed out of your Pfibong.
There's still an inhaled insulin product on the market, Afrezza, which looks like a weird vape.
manulcapital retweeted
So insightful it's obvious in retrospect. teapaw.org/curtis_yarvin/status/2…
dont worry It's temporary
It's fairly easy to make a fake seinfeld clip that tricks 50% of people. Harder to trick 80% of people, extremely hard to trick 99% of people
We have an assumption though that this will nonetheless happen. At some point, the technology will improve enough that tricking virtually 100% of people will be effortless. It's probably true
But there's a chance that instead, people decide it takes too many resources and isn't worth it to progress past an arbitrary point. Maybe they decide tricking 98% is good enough and more on.
This happens a lot in society. You can break almost every law without getting caught if you're top 2% determined/resourceful. It's often not cost-effective to make something robust against the extreme tail edgecases
In the case of AI video this would result in there being a group of people, the 2% with the highest discernment, being trapped in some cartoonish hell where everyone in their life is constantly sitting there laughing at these AI generated videos, sharing AI generated news, forwarding AI generated reels with them saying things like "haha, this is so true, check it out man!"
As they sit there slowly going insane, knowing that this is all totally fake but being unable to prove it to anyone
AI is purposely nerfed at writing so that people like this don't have a mental breakdown btw
You could trivially make a superhuman writing AI, but people don't bother
Jesse Eisenberg says he sometimes "experiments with AI" to see if it's good at writing:
"I’m totally paranoid. I sometimes experiment with AI to see whether it’s good at writing. And it’s still not. I’m paranoid, so I check and it still isn’t. Thank God. I say, 'Write a scene like this.' I’m just testing it to make sure it’s bad, and it’s still bad. So I’m relieved, but, like, how far away is it from being good? How far are we from a kid putting on a pair of goggles and saying, 'I want to watch a movie about me where I rob a bank and then kiss the girl I’m in love with at school. Here are pictures of her, here are pictures of me' and then they press play? I’m paranoid about that."
variety.com/2026/film/awards…
