Daigo Kusunoki

This resonates a lot. I started SOUNDRAW because I personally couldn't find the right music as a dancer, the problem was mine before it was anyone else's. That's the only reason I think it worked where other attempts in this space didn't. Founders who are solving someone else's problem they read about are usually a step behind founders who lived it.

The "AI-native agencies" idea is interesting to me because music is one of the last creative services that hasn't been restructured around AI tooling yet. Most agencies / businesses still license from the same catalogs they used 10 years ago. We've been building infrastructure for that shift since 2020, the demand is real, the licensing just needed to catch up.

I'm about to launch for the first time so I've been thinking about this a lot. What worked when I talked to other makers: stop optimizing for upvotes and optimize for showing up as a real person beforehand. The platform has changed but people backing people they actually know hasn't. We're launching our API on the 24th and honestly the conversations in threads like this one have been more...

Late to the party, but this is something we deal with daily at SOUNDRAW. The biggest trap we see developers fall into is assuming "AI-generated = royalty-free." It's not, the copyright situation depends entirely on what training data was used and how the platform licenses outputs. We made the decision early to build commercial licensing into our API at every tier. Partly because it's the right...