This is the first public release of Hallucinote, an integration of Claude Code and Ableton Live. Hallucinote helps with the entire lifecycle of songwriting, including ideation, instrumentation, harmony, rhythm (including microtiming), arrangement, mixing, and production.

Hallucinote is focused on assisting songwriting, not doing it for you. Key features:

- Git-first representation of songs, using Ableton as an audio rendering and performance capture surface

- A decision log for each song, covering all aspects of songwriting and music production

- Composition skills adhere to the song’s decision log and write with micro-timing and other niceties

- Support for meter changes, polymeter, and alternate tunings

- Skills to critique melody, harmony, and rhythm against your stated goals (NOT “is this a good melody in the abstract”)

- Round trip: perform MIDI tracks in Live, pull down into Hallucinote for embellishment, push back to Live (no audio track round trips yet)

- Captures every track and bus audio, with LUFS and other key measures Live-side [requires M4L / Suite]

- Mix analysis done by Claude code, pulling in external audio analysis tools as needed to identify clipping, muddy frequencies, misbehaving effects [requires M4L / Suite]

- Automation, sidechains, device chains of arbitrary depth, envelopes everywhere Live supports them

- And yes, an MCP server with >100 actions

I’m REALLY trying to avoid making an AI music generator; this isn't Suno. I wanted an expert assistant who’s willing to help with any aspect of music creation, who will call me out if I'm not thinking of something like “where are the vocals going to fit in this mix”, and who will handle tedious stuff like “hey let’s try a few ways to get from D minor to F# Phyrgian over the course of the break” or "let's try a mix that doubles the first chorus and simplifies the drums in the first half"

I'd love feedback... it's been a hobby project for a while and I figured it's time to share.

Comments URL: https://news.ycombinator.com/item?id=49271123

Points: 3

# Comments: 0

So instead of AI that pretty much everyone can use, we could have had 3 or 4 weeks of business as usual?

There are 6 different transfer speeds, from 480mbps to 120gbps.

There’s DisplayPort yes/no.

There are 4 thunderbolt genrations, each with its own marker.

Some of those are dependent (480mbps cables don’t have markers), so it’s not a full permutation problem. Only about 20 different combinations are valid.

Is the proposal to have 20 different physical connectors? Or to prune the space down to 6 or so by requiring all cables to be very thick, or to support higher transfer speeds than make any sense for a charging cable?

It’s not a leaky abstraction. It’s a physical interface standard that does not guarantee electrical properties. Same way you can plug a 6A cord and a 20A cord into the same electrical outlet.

There’s a good argument for using the funds to support mental health programs, but as far as paying for damage done, Meta is getting off very light here. Classic privatize profits, socialize costs.

Huh? That’s quite a whopper to drop as if it’s an obvious truth.

It’s good to diversify supply chains. Eggs-in-one-basket is bad for many reasons, even if it’s your own domestic basket.

Trade interdependence is good for competition, reduces risks from disasters, and creates disincentives for war.

And the market apparently thinks offshore wind is a good investement for the US, to the point where the government has to pay companies not to do it.

Only difference is I’m prompting LLMs rather than engineering teams. But honestly it’s not all that different, except that in the hobby flavor I can write specs that would take tens or hundreds of engineers to implement.

Maybe it all comes down to semantics: does “programming” mean typing in code, designing app architecture, or producing software? When I was young I was probably 75 / 5 / 20 between those. Now I’m 5 / 20 / 75. Maybe it merits a different word?

But I think you’re on to something important: framing programming as a competition of intrinsic skill where rules are necessary to equalize. That’s certainly true in some cases, and more power to those who enjoy it.

I guess I’m just more of a product guy these days, where the equalizer is the market and the consumers rather than the implementation methodology.

Few cars are manual steering, manual brakes, manual transmission.

Few grappling bodies go all Greek-naked.

Few chess bodies forbid researching previous games or your opponent.

The list is indeed endless, and every single person thinks they are using just the right amount of assistance; anything less is pointlessly difficult, anything more is cheating that takes the sport out.

Me, I've been a hobbyist programmer since youth (6502 assembly, represent). I'm having a blast "writing" software without touching a line of code. Sure, I'm old, and maybe I'm like those guys blasting Veyrons through highways, relying on traction control and all the computers to let me drive beyond my abilities. And... ok? I'm having fun, even if I'm not doing it right, by random anonymous peoples' views.

Let’s say I ask “who created Linux?” Claude correctly tells me Linus Torvalds, and links to Wikipedia.

There are probably thousands, maybe hundreds of thousands of other sites that have that some piece of information. Are LLMs supposed to link to every site?

Most sites do not have unique information at all, and even sites that do rarely contain only unique information. Implying that most sites deserve links because they were crawled seems like a statistical fallacy. You could say the same thing about the percent of sites crawled by Google versus ever showing up on first page of results.