Reimagining Ownership for Networks and AI

Originally published on Substack — Aug 2025

On April 12, 2025, Jack Dorsey tweeted, “delete all IP law.” Elon Musk immediately replied, “I agree.” The backlash was instant. Critics warned this would gut protections for artists and creators; others cheered the prospect of finally freeing innovation from the dead hand of gatekeepers.

It stayed with me. Not because I fully agreed, but because the comment forced me to pause. To look back at my own decade spent working on property rights in the digital realm, and to ask again: what is it we actually want from property rights in the age of networks, code, and AI?

The Tension at the Frontier

Part of me respects people like Dorsey and Musk merely because they operate at the edge of the frontier. They see the weaknesses in the system at a scale I never will. More often than not, I think they’re right. But there’s another part of me that feels a sting of guilt. I’ve spent 10 years trying to work on this exact issue: how to make data and other digital assets fit into a property rights framework. And I can’t help but feel I may have gotten the whole direction wrong.

In the early days of the internet, lawyers defaulted to treating everything as IP. Copyrights, licenses, restrictions: that became the language of digital ownership. Looking back, I think that was a mistake. It bolted analog constraints like territorial borders and slow enforcement onto a borderless medium built for speed and flow. But even as I tried to build alternatives with Bitmark, I still found myself falling back into those same patterns.

So when I read Dorsey’s words, I asked myself: why would you delete something? You delete something either because it’s no longer needed, or because you need to test whether it is. Musk has a rule: delete so much that you end up putting back at least 10% of what you cut. The point is to expose what’s truly necessary. And if you apply that to IP law, you’re left with a stark question: what parts of property rights do we still need? Which parts are foundational, and which parts have become deadweight?

Lessons from Bitmark

When I launched Bitmark, I thought patents would be our foundation. In 2014 and 2015, I spent enormous time and money filing them. The plan was to secure a wide set of claims and then either open-source them or use them defensively. It felt logical at the time. But the reality was brutal: each revision cost thousands of dollars, and we had to file in multiple jurisdictions just to be credible. Even then, there were gaps, countries where we had no protection at all.

I never paused to ask the simplest question: why are we creating new property inside a legacy system, if the whole point was to build the first borderless property system? We should have dogfooded our own vision.

The patent process also revealed something more nuanced about innovation. Similar filings started appearing all around us. Not because anyone copied us, but because the ideas were in the air. People everywhere were starting to imagine blockchains recording more than money. In theory, that should have been validating. In practice, it was maddening. And if I’m honest, when I finally realized patents weren’t the right foundation for Bitmark, trying to bolt a borderless digital system onto analog, territorial laws that contradicted our vision of code-enforced ownership, I didn’t feel relief. I felt guilt. I had wasted so much time, money, and energy chasing a dead end that pulled us away from dogfooding the very freedom we were building.

Why IP Law Misaligns with Innovation

That guilt deepened as I thought back to earlier chapters in my career. In the Linux world, I saw the SCO lawsuit, a case so absurd it almost derailed one of the greatest community-driven technologies ever built. Later, working on mobile phones, I discovered that Qualcomm captured 15% of the landed cost of every device, not on profits, but on the entire value of the phone, all because of radio patents. Meanwhile, the real breakthroughs were happening in the application processor and the software running on it. The asymmetry was staggering.

And then there were the endless Apple versus Samsung lawsuits. Watching giants sue each other over rectangles with rounded corners was exhausting. For hackers and builders, IP law feels less like a framework for innovation and more like a global mutex: anyone can lock it, and everyone else has to wait.

The Limits of Private Ordering

The alternative to state-backed IP is private ordering, rules crafted by non-government players, outside the reach of borders or bureaucrats. Think back to feudal Europe before the Peace of Westphalia in 1648, where guilds and merchants self-regulated trade without a central king calling the shots, or indigenous customary laws that evolved from community needs rather than top-down decrees. It's a concept anarchists dream about: societies organizing their own norms, bottom-up.

In our digital world, this shows up everywhere, from company bylaws to the fine print in your social media Terms of Service. We've traded this for seamless experiences, scrolling without friction, apps at our fingertips, but at a steep hidden cost: massive centralization of power in unelected hands. Apple's App Store isn't just a marketplace. It's a private legal empire dictating what innovations live or die, often burying arbitration clauses that strip away real choice. Can you truly "agree" when disagreeing means getting locked out of the conversation? This variant of property rights, controlling access and ownership through code and contracts, threatens human autonomy on a global scale, shifting us from owners to renters in our own digital lives.

Yet here's the flip: We can reclaim it. By blending blockchain's borderless enforcement with smart contracts, we create private orders that decentralize power instead of hoarding it. The internet, as our collective nervous system, deserves its own self-governing framework, one where individuals, not intermediaries, hold the reins.

Transaction Costs and Blockchains

Private ordering sounds liberating, rules by us, for us, but in Big Tech's hands, it's often a velvet trap, hiking hidden costs like lost privacy and forced consent. This is where Ronald Coase's insights cut through the fog. In one of his landmark ideas, he explained why firms exist at all: Markets are expensive, riddled with transaction costs, searching for partners, drafting contracts, enforcing trust, resolving disputes. Those frictions decide whether we build in-house or outsource to the wild.

Blockchains are revolutionary here because they slash those costs to near zero, encoding agreements in code that runs automatically, borderlessly, and without middlemen. But Coase's deeper gem, and the one that really unlocks property rights, is his take on externalities: The messy spillovers, like a neighbor's noise or a factory's pollution, that markets ignore. His theorem boils down to this: If transaction costs are low and property rights are clearly assigned, people can bargain their way to fair, efficient outcomes, no matter who starts with the upper hand.

In the digital realm, externalities are everywhere, your data trails exploited without credit, creative works remixed without consent. Blockchains don't just cheapen deals. They let us assign real ownership to those intangibles, turning them into tradable assets. It's like giving every drop in a shared aquifer its own deed—suddenly, waste turns to stewardship, and individuals gain the power to negotiate on equal footing. That's the uplift we need, not more gatekeepers, but tools that empower autonomy at scale.

Government at the Merkle Root

Coase's theorem hints at a balanced path: With low transaction costs and clear property rights, we can negotiate our way out of externalities without heavy-handed oversight. But that doesn't mean ditching governments wholesale, far from it. We still need their stability to anchor the basics: Rule of law for security, predictability, and legitimacy when code alone can't handle the messiest human edges.

Picture it like a Merkle tree in a blockchain: Government sits at the root, providing that unshakeable foundation. From there, the branches fan out through private ordering, smart contracts, and community governance, evolving freely without the state poking into every leaf. It's not deletion. It's delegation, keeping the core solid while unleashing the edges.

I'm eyeing jurisdictions like Wyoming, with their DAO-friendly laws that blend state backing with digital flexibility. At Feral File, we're not rushing in yet, not out of caution, but to first distill Bitmark's hard-won lessons into rocket fuel. Bitmark was theoretical at its core: Dreaming up borderless property systems, chasing patents as placeholders. Feral File flips that to applied, real ownership baked into daily art experiences, where low transaction costs mean global access without gatekeepers. That shift from abstract blueprints to tangible impact? It's the tuition paying off, turning past detours into a clearer shot at uplifting creativity on a massive scale.

Programmable Property Rights

So how do we actually stitch this into something workable? One way I picture it is to treat ownership less like a static deed and more like a living graph that evolves with every contribution and remix. Start with ERC-721 tokens to mint unique digital assets—art, code, writing—anchoring provenance. Add ERC-6551 token-bound accounts, which let each token function as its own smart wallet. Suddenly a token isn’t just a receipt; it can hold sub-assets, sign transactions, even own other contracts. A creation can grow with its ecosystem instead of sitting frozen on a shelf.

To keep track of who built what, I imagine contribution DAGs, directed acyclic graphs, something like Git’s commit history, mapping inputs over time. Weights could reflect effort, impact, or even community judgment. Those graphs then feed into dynamic royalty splitters: smart contracts that route payments automatically based on the recorded contributions. As the graph updates, so does the split. In theory, you don’t need lawyers to haggle over percentages; the system itself carries the logic.

The harder part is what to do with the gray areas. This is where AI might become useful as an oracle. Large models are surprisingly good at parsing intent, something closer to what courts do today, just slower and at much higher cost. An AI oracle could act like a digital arbiter: ask, “Is this a fair derivative of the original work?” and it returns a ruling with an explanation, feeding directly into the contract. Done right, that could seed a kind of internet-scale arbitration layer; cheap, auditable, and open to appeal or multi-oracle consensus. Not a perfect solution, but maybe enough to lower the “justice tax” that keeps most creators from defending their rights at all.

Here’s a sketch for a simple splitter oracle:

// Pseudocode for a Contribution-Based Royalty Splitter with AI Oracle

contract DynamicRoyaltySplitter {
    mapping(address => uint256) public contributions; // DAG-weighted shares
    address public aiOracle; // LLM-based judge for disputes

    function addContribution(address contributor, uint256 weight) external {
        // Update DAG: Verify via oracle if weight is fair
        (bool isValid, string memory reasoning) = aiOracle.query("Is this contribution weight accurate based on history?");
        require(isValid, reasoning);
        contributions[contributor] += weight;
    }

    function distributeRoyalties(uint256 amount) external {
        uint256 totalWeight = getTotalDAGWeight(); // Sum from graph
        for (address contributor in contributors) {
            uint256 share = (contributions[contributor] * amount) / totalWeight;
            payable(contributor).transfer(share);
        }
    }

    function resolveDispute(string memory query) external returns (string memory) {
        // AI oracle parses intent: e.g., "Was this a derivative? Who gets attribution?"
        return aiOracle.judge(query); // Returns ruling + explanation
    }
}

This isn’t a blueprint. It’s more of a hack to test assumptions. The appeal is obvious: for the price of a coffee, you could resolve disputes that today would take months in court, and maybe unlock new forms of collaboration that traditional IP systems usually smother. But the risks are just as real. If the underlying models are biased or captured, the whole system is compromised, like poisoning the well. The only way it stands a chance is if it’s built to be open, auditable, and diverse. Then maybe we start to see ownership shift away from scarcity games toward something that actually supports shared progress.

Limits and Open Questions

Bitmark’s detours into health data and music weren’t failures so much as reconnaissance. Health was tangled in regulations that prioritized compliance over flow. Music was too consolidated, locked down by a handful of rights-holders that turned creativity into a toll road. Those efforts taught me where new systems are least likely to take root.

The more promising ground is in domains that are global, lightly regulated, and not yet dominated by entrenched players. There, programmable property could move from theory to practice: reducing transaction costs enough that ownership starts to feel natural, not burdensome. Of course, markets can twist toward speculation and quick flips, but that risk comes with any new system. The opportunity is to design things so people actually use and live with what they create, instead of hoarding it away. It reminds me of the early days of open source: messy and full of bugs, but transparent enough that people could dive in, fix things, and slowly turn chaos into Linux.

The bigger opportunity is in the unsolved edges. These aren’t dealbreakers; they’re invitations. How do we keep contribution graphs honest without introducing gatekeepers? (Maybe zero-knowledge proofs could verify effort without exposing everything.) How do we balance transparency with privacy, keeping systems auditable without exposing individuals? And if AI oracles are making judgments, how do we ensure their decisions are explainable and open to appeal? These are not answered yet, but that’s exactly what makes them worth exploring.

Passing the Baton

The right calibration for me now is lessons compounded, not burdens carried. Bitmark aimed high, sketching big theories of borderless rights, but we did not ground-test enough. What I am chasing now are applied edges: ways to encode ownership so barriers fall away, and abstract claims turn into tools people can actually use. The goal is straightforward: lower the cost of creating and sharing, and resolve disputes quickly and fairly, without borders getting in the way.

For the next builders, the trap to avoid is binary thinking. Do not throw out centuries of property wisdom. The first principles, such as clear assignment of rights, fair bargaining, and some rule of law at the foundation, still work. What is broken is the machinery around them: bloated registration, uneven enforcement, systems that no longer serve their purpose. Those are the parts to rethink.

The path forward looks more like evolution than revolution. Small, feedback-driven changes that compound into something bigger. If done well, property rights adapt to networks and AI without losing the human core. That is where the promise lies. Not in producing more, but in creating more autonomy, where people can shape their worlds without gatekeepers holding them back.

Final Reflection

Bitmark’s deepest lesson was the gap between telling and showing. I told stories about programmable rights, built APIs and scaffolding, but we never gave people that direct “I get it” moment. The kind where you hold something in your hands and it makes sense without a single word of explanation. That is my new north star: build experiences where the vision is obvious in use, not in description.

This reflection is bittersweet. I feel guilt for the detours, but also gratitude for the ways they shaped the ideas. I feel urgency because the stakes are real: human autonomy, the ability to create freely, and the chance to collaborate without barriers. At the same time, I feel hope because the tools now exist. Blockchains reduce costs, AI can help us parse intent, and decentralized governance can turn groups into stewards rather than gatekeepers. These are not gadgets. They are raw materials we can shape into systems of abundance, where ownership shifts away from scarcity games and toward shared progress.

This is not a manifesto or a pitch. It is closure, and also a hand-off. To those who keep building: do not try to fit tomorrow’s ideas into yesterday’s boxes. Start from how collaboration actually works today. Ask what ownership should look like in that reality, then test it until it holds.