Every AI lab’s legal team has the same internal slide deck. Rights holders’ legal teams have the same one too. Neither has read the other’s. That’s the entire industry, in one sentence.
I build markets where there
weren’t any —
for IP, for data, for creators.
Founder of IPTO. Currently rebuilding the licensing layer underneath AI. Earlier: Create Protocol, Neuron Gym, Ojam, Wild Monks. IIT Roorkee. LA-based, made everywhere.
building the licensing layer for AIWhat I noticed this month
Short, dated observations. The notice-it-first muscle, in public. All notes →
A startup whose moat depends on a regulator moving fast is not a startup. It’s a lobbying campaign with a cap table.
Notice how every AI ‘ethics’ conversation collapses the moment payment is mentioned. Ethics is the abstract noun for a concrete refusal to pay.
In Tokyo, the meeting that matters happens after the meeting. In LA, the meeting that matters never has a meeting. India runs the meeting twice and uses the second one to argue.
Build markets where there weren’t any.
The thing in front of me
Not a manifesto. Just where the attention is. Full /now page →
Building
IPTO — pushing the licensing API toward general availability and signing the first cohort of rights holders.
Thinking
Why every AI policy debate ends up at the same place: the absence of a real-time, machine-readable rights market.
Making
A short Abhi Cris record — built around tempo as a single argument across six tracks.
Open to
Talking to rights holders, AI labs, and policy people who actually want a market, not a moratorium.
Five claims I’d defend in public
The arguments behind the work. Each one is a single sentence with a longer essay behind it. All ideas →
№ 01
Licensing is the missing market
AI didn't break copyright. It revealed that the real-time, machine-readable licensing market never existed.
Every AI lawsuit is a market price discovery in the most expensive way possible — through litigation. The fix isn't tighter copyright; it's an API where rights, terms, and payments resolve per call. That's the bet behind IPTO.
№ 02Data is property, not privacy
Privacy law is the wrong layer. Personal data needs the legal grammar of property: ownership, transfer, consent, royalty.
Privacy frameworks try to hide data; property frameworks let people own it, sell it, withdraw it. The latter is what large models actually need to be legal at scale, and it's what humans actually need to be paid.
№ 03After AGI, taste is the only moat
When generation is free, the scarce inputs are taste, provenance, and intent.
We've spent twenty years compressing distribution. The next twenty will compress generation. What remains expensive — the bit machines can't fake — is judgment about what to make and why. Networks should pay for that signal.
№ 04Every track is a tiny operating system
Composition is design under the deepest constraint a builder can take on: time you can't pause.
Tempo, key, and texture are state. Surprise is the only metric. Releasing music as Abhi Cris is the same instinct as shipping software — finite resources, irreversible commits, an audience that can quit any second.
Edits beat additions. Cuts beat features. Constraint beats budget.
Public bets, with dates
I take stances. Every bet has a resolution date and a confidence number — so when it’s wrong, you’ll know. All wagers →
-
By end of 2027, at least one major AI lab will route inference through a programmatic licensing protocol — not as PR, as architecture.
Litigation cost is now exceeding what a per-call licensing fee would have been. The first lab to admit it publicly wins the policy narrative.
-
A meaningful US bill — meaning one that gets a committee vote — will frame personal data as a property right, not a privacy right, before the end of 2026.
Privacy frameworks have failed to produce payable cash flow. The legislative bottleneck is who collects, not who hides.
-
Two of the five highest-funded ‘creator economy’ networks of the 2021–2023 vintage will sunset or quietly pivot off creators by mid-2027.
Distribution-only networks were always renting their growth from another network. Generation getting cheap removes their last input cost moat.
The questions everything else hangs from
Use these as doors. Each thread links the ventures, essays, and music that orbit it.
IP economics
Licensing as the missing market for the AI era.
AI didn't break copyright; it revealed that the licensing market never existed at the right resolution. The fix is API-native, machine-readable, per-call rights — IPTO's bet.
Post-AGI creativity
What humans make when machines can make anything.
When generation is free, taste, provenance, and intent become the scarce inputs. Creative networks should reward the upstream signal, not the downstream byte.
Human data rights
Treating personal data the way we treat property.
Privacy frameworks fight the wrong war. Rights, ownership, and consent are property concepts — and that's the legal grammar AI needs.
Decentralized creative economy
Networks that pay creators by default, not by lawsuit.
Royalty rails, attribution, and incentives belong in protocol, not in legal departments. Create Protocol is the long bet on this stack.
Operator's craft
Building from zero across India, Japan, the Gulf, and the US.
Bootstrapping teaches the unit economics that decks hide. Capital is fuel; clarity, taste, and timing are the engine.
Music as system
Composition is design under deep constraint.
A track is a tiny operating system: tempo, key, and texture as state, surprise as the only metric. The same instinct ships software.
Consciousness & tech
Where the inside of the head meets the outside of the network.
From Neuron Gym to AI: the most interesting interfaces are the ones that change the user's interior, not just the screen.
Notice everything. Commit to one thing. Repeat.
What I’m building, what I built
IPTO is the present tense. Create Protocol and Human Data Rights are the long bets. Earlier ventures left the infrastructure today’s work runs on. Full ventures + experience →
IPTO
CURRENT FOCUS · AI-native licensing infrastructureGrowing IPTO — an AI-native licensing platform so developers and creators access ethical, licensed IP through APIs. Bridging the rights gap in AI and building infrastructure for the next generation of creative technology. See ipto.ai and this hub.
-
Create Protocol
Web3 / AI InfrastructureDeAI network accelerating human creativity and intelligence for a post-AGI world — decentralized creative economy infrastructure.
-
Human Data Rights
Data rights / policy + toolingA founder-led initiative exploring practical frameworks for human data rights in the age of AI.
Essays I’ve actually finished
On IP economics, post-AGI creativity, human data, and the operator’s craft. All writings → · RSS
-
Licensing is the missing market
AI didn't break copyright. It revealed that the real-time, machine-readable licensing market never existed.
2026-04-22
-
Data is property, not privacy
Privacy law is the wrong layer. Personal data needs the legal grammar of property: ownership, transfer, consent, royalty.
2026-04-08
-
After AGI, taste is the only moat
When generation is free, the scarce inputs are taste, provenance, and intent.
2026-03-18
-
Every track is a tiny operating system
Composition is design under the deepest constraint a builder can take on: time you can't pause.
2026-02-26
-
Decks lie. Unit economics don't.
Fifteen years of building between India, Japan, the Gulf, and the US. The only artifact that ages well is the P&L.
2026-01-29
Capital is fuel. Clarity, taste, and timing are the engine.
The music isn’t a side project
Every track is a tiny operating system. Releasing as Abhi Cris — electronic, ambient, experimental. Music page →
album Abhi Cris on Spotify Full discography under the Abhi Cris alias — electronic, ambient, and experimental. The canonical home for finished work. sketch Ambient Exploration 2024 Eight-and-a-half minutes in C minor at 60 BPM — a single piano line slowly accreting synthesizers, reverb, and field recordings until the silence is louder than the chord. live Wild Monks era 2017–2020 Pre-Abhi Cris years — writing, composing, and distributing music globally. Where the live instinct came from before it became electronic.
What fifteen years actually taught me
- Bootstrapping is a research method. The P&L is the lab notebook — every line is a hypothesis under test.
- Operating across India, Japan, and the US trains one muscle above all: reading what the other room is willing to say out loud.
- Creator-first is not a marketing posture; it's an architectural decision about who the system pays first.
- The reason the AI rights debate keeps failing is that it's being argued in copyright when it should be argued in market design.
Audiences quit any second. Design assuming they will.
If any of this points at you, find me.
Rights holder ready to ship a programmatic licensing pilot? AI lab tired of litigating instead of paying? Policy person who actually wants a market? Builder at this intersection? Yes, please.