law for good: deflate to proliferate

This article will discuss the various ⚖️/acc innovations our team has done in 2024. It’s written in part to fulfill our artifact requirement as a Devcon 7 Legal Scholar. Thank you very much Ethereum Foundation for the honor.

This article is mostly focused on my work during 2024, but please visit here to learn more about the work our firm Ixian.tech did over the year.

We’ve done a lot in 2024, and did a lot of cross-pollination with network states.

Talks from conferences this year shared here , here and here among other places.

We refined our concept of “law is a public good” to “law for public good” to now just “law for good.” This keeps us as far away from public goods myopia as possible, and ensures we live by the spirit of goodness throughout the law.

Perhaps my most important insight this year is how long the legal industry has been an inflationary cost and how, comparing its potential technological development to the transportation industry’s historical development, I have confidence that if we continue to apply engineering principles to the legal industry, we can invert legal costs in most industries from inflationary to deflationary.

Given how legal industry costs are costs included in all industries, lowering legal costs will offer outsized prosocial impact in all sectors. Perhaps even more so than energy cost savings because, unlike energy, legal abundance naturally compounds on itself. Remember, fiat money is a sophisticated legal instrument -- IP-backed promissory note, no gold behind it -- so compound interest is a direct aspect of my contemplations; we compound more legal quanta than just interest.

Throughout the ideas I share here, the common thread is “law for good: deflate to proliferate” the law.

types of law

law as physics

The laws of physics, developed over time, are clearly deflationary. Maxwell equations, Feynman Infinities and Newton-Einstein inertia. They give us the foresight to anticipate prosperity, and enable us to compound on prosocial complexity.

Code-is-law is interesting from this perspective. But the immutability of physical laws is that is, unlike code-is-law. Practically, that means the laws of physics are expressions of unbreakable consistency within all observations of a phenomenon whereas code-is-law are just logical expressions being run by many machines in the same way with deterministic outcome. Deterministic because is AI code-is-law? Probably not.

Still very helpful to think of code-is-law as a physics engine when immutable. Physics, from this perspective, is about laws that are immutable in perpetuity. Of course, the immutability of laws on Ethereum is not as immutable as the laws of physics, but within most jurisprudential time scales, they are essentially the same.

My experience developing the (c-b)/a base-scale informs my understanding of physics, as I had to observe geometry with an electromagnetic calculus to guide my discovery, that did include a bit of design. Shown here below as a method to tessellate gradient descent that uses the same energy as foreshortening techniques but within the 0 to 1 logarithmic bases unit circle rotating along π/2 dimensions.

law as the software of society

What scope does the software run on? For lesser scopes, including peer-to-peer, is there a kernel or framework we can use to streamline globally? Can we test new regulations in running code opt-in or free city sandboxes first, rather than all-at-once nation-state deploy untested code?

Of course, if consumer goods included at least some servers, not just clients, we’d be much more capable at manifesting running code opt-in; good thing AI agents are finally bringing this need home.

Anyway, deploying a dispute resolution system like Kleros on a nation state is not nearly as derisked as doing it as-is, where parties opt into Kleros as dispute resolution. Indeed, some sovereign courts of appeal are not just recognizing but embracing Kleros as a legitimate form of dispute resolution within their terra firma legal systems (source).

jurisprudential inventors

This industry -- legal engineering -- is fledgling yet already mighty. Just 2 legal engineers have created over $20 trillion USD in value, changing the fabric of society.

Legal engineers recognize that law is a mechanizable industry, and as inventors, we can compound industrious novelty upon industrial novelty and develop deflationary technological feedback loops in order to recognize law, not as presently over 150% civilization-destroying inflationary, but as civilization-creating deflationary.

My main teaching against compared to traditional lawyers is that, to me, program language is not separate from natural language but, rather, an evolution of it. Therefore, not being literate in programming language is not being literate in natural language as it now stands. As wordsmiths who demand a lot of precision from language -- hence the development of legalese -- it’s unfortunate that the vast majority of us remain illiterate, since the impetus is improving accuracy.

Taking the realization inflationary legal costs destroy civilizations into account, to be honest, the legal community’s illiteracy is not just unfortunate, it’s an existential threat to civilization. Taking the speed of AI development into account, inflationary legal costs at this moment in time is actually a species-level existential treat.

Those of us lawyers who are literate, have found some very low-hanging fruit that has already substantially improved the law by trillions of dollars (Goldfarb & Szabo). With the advent of AI co-pilots, hopefully we can upgrade the rest of the legal community back to literate within just a handful of years. If Sequoyah could do it, so can we!

law is inflationary

A narrative from our venture studio, based on bureau of labor statistics numbers:

The idea is that legal vehicles can achieve similar levels of deflationary transportation / transaction cost as transport vehicles have enjoyed: for the same reason, mechanization. Automation is also important, but giving up full cybernetic control will likely be achieved after the hyperspeed environment manifestation has matured. Remember, from a horse-and-buggy perspective, cars and freeways are hyperspeed. And we don’t yet have self-driving cars other than in a very limited number of locations.

Próspera has paved the way, and Zuzala has too

This 2024 article explains our cosmolocal sovereign-tech vision.

local sovereign-tech, especially terra firma

Establishing our law firm a year ago, the freedom of operating in the Próspera free zone with a legal engineering inventor’s creativity, has already yielded valuable experiments as well as ways of seeing how the law can be more prosocial.

The polycentric nature of Próspera has also proven very helpful in developing new legal frameworks because we don’t feel so much pressure to get the solution 100% correct the first time. Instead, we move forward with a solution we think might be a good idea, and compare it to the myriad of other solutions we are testing out. This, then, allows us to develop a dialectic, based on empirical developments, not possible in any other jurisdiction. We find we’re naturally designing regulation much more like how developers design code, gravitating to modular structures so we can test different subsets, such as the health regulations we’ve worked on.

It’s unfortunate the Honduran Supreme Court decided Próspera has been unconstitutional from the start. One of the silver-linings from this setback, however, has been us realizing that we can better protect assets and help achieve network state aspirations by having free cities sign multilateral agreements with each other. More about this in the “refugee network” section.

Zuzalu community events are also local, and their localization is extremely valuable. Unlike Próspera and other free cities, however, Zuzalu communities aren’t relying on a host sovereign to develop sovereign-tech. Instead, they are developing the sovereign-tech on the cosmo level, via the world computer described below. If absolutely necessary, it’s possible that these Zuzalu communities could stay entirely cosmo. But, it is so much easier to coordinate with and bond with other projects by travelling the world with them.

And, besides being fun, this cosmopolitan caravan strategy helps progress the practical worldwide physical remoteness necessary for this sui generis form of sovereignty to emerge in the first place.

cosmo sovereign-tech, especially world computer

It is arguable that the cosmo form of sovereign-tech is sui generis sovereignty. Regardless, world computers such as EVM systems -- where multiple machines distributed around the world share a global-operating-system state together -- are recent technologies.

Sui generis has been used to describe new forms of property, and, as such, we’ve used it to describe the digital asset parallel financial universe forming on these world computers, but analyzing sovereignty as potentially sui generis is not something we’ve seen before.

Thinking about the laws of physics, it makes sense to recognize this as a new form of sovereignty. Traditional sovereignty is largely determined by the laws of physics. Or, in the very least, sovereignty is the area of law that cannot artificially circumvent the laws of physics because it’s the area of law that’s not a subset of a wider legal environment. Might-makes-right and other last appeal power dynamics determine sovereign-to-sovereign disputes.

So, the fact that no sovereign can physically shut down these world computers -- and in the past no technology like this existed -- means that these world computers do have a heretofore nonexistent form of sovereignty.

This sovereignty exists in finance law. And we’ve figured it out for some other basic legal complexities in finance, such as escrow. Given how so much of dispute law uses financial remedies, monies are far preferred to injunctions, having this layer available already gives us the ability to cosmo sui generis mechanize a substantial portion of some of the more impactful areas of law.

It’s important to note, though, that as natural persons, we always embody the local. Thus, there’s always the possibility you could be called to terra firma jurisdiction, and you cannot fully rely on this sui generis form of sovereignty. Habeas corpus remains a critical right to uphold locally, no matter how cosmo we get.

Have a look at some of our early slides on the topic, during ZuVillage Georgia this Summer.

airplane registration example

This is the first project where we thought about combining the public goods nature of regulation with for-profit motivations. In this case, our client wanted to register an airplane in the jurisdiction, and since he was the first to do so, he also had to create the regulation. The regulation is a public good. To encourage him spending on developing this public good, we also offered to collaborate on making a product with him on top of the regulation. The product is for-profit and designed to be as retail-friendly as possible.

In other words, we avoided this never-manifesting, quiet but very prevalent, tragedy of the commons by packaging the commons as a loss-leader body on which we attached a profit motive header.

In a polycentric environment, the private sector development of loss-leading regulation through the development of for-profit products could encourage the development of highly efficient and user-friendly regulatory systems. There’s no private capture of the means of production. Because, if the market is healthy, different means should be in continued innovation unless close to maximum efficiency. For most areas of governance and law -- the legal engineering industry -- we’re not close to maximum efficiency.

This airplane registration project is we started working on our venture studio project management system. Though we’re still learning the ropes, this process has been very helpful in systematizing an otherwise highly chaotic and uncertain discipline.

Refugee Network

We first shared this presentation at EdgeCity Lana, 2024 on a panel with Primavera during Network Nations week. The idea stemmed from when the Honduran Supreme Court this Fall deemed ZEDEs unconstitutional from the start, and the conversations we had with other free cities immediately thereafter brainstorming together on how they could help us.

The most important slides are our IF/THEN recommendations:

During our panel, we realized how important it is for these local nodes to record all of their legal information within cosmo storage, so that a host sovereign cannot threaten the destruction of knowledge about extant and legacy legal relationships.

cosmo-precedent for the ethereal ages

inverted-precedent

This idea was funded by the round #3 public good fund tokens that Optimism gave to LexDAO.

I was secret-vote impeached by LexDAO because of my commitment to not wasting Optimism’s funds and building out this product. At least, that we don’t sabotaging the prototype experimentation stage, whether or not we decide after that to scrap the prototype rather than continue forward along the 0 to 1 path.

Luckily, we’ve found more success in the Zuzalu community than LexDAO, and we now have a scope of work ready to deliver during Q1 2025. This will be done by RaidGuild and is an upgrade to their already very capable y33t DAO system. Details in the EthSign signed Memorandum of Understanding “MoU” with Ethereum Community Fund here:

One of the most exciting things about this y33t DAO inverted-precedent -- besides teaching laypeople how to think more judiciously -- is that a project is held accountable throughout the project’s runtime. So far, the only grants mechanism that operates during the runtime is conviction voting. With the y33t DAO inverted-precedent, if a QF vonater decides to ragequit, with a precedent trained AI co-piloted opinion published as to why, they return the matched funds to the QF pool as well as get the remaining percentage of their own vonation back.

Spending proposals from the y33t DAO inverted-precedent also use precedent trained AI co-piloted opinions about why they are making the proposal, with the added complexity that concurring and opposing opinions can be part of the proposal, and in some cases this expression of diverse reasoning systems is encouraged.

One recognition here is that, though I’m trying to offer alternatives to the state-coerced adversarial disputes that are our existing judicial system, some elements of this system are some of the most sophisticated coordination tools in society. Namely, when working well, the opinions of the Supreme Court is how we intelligently synthesize society’s most adversarial and unresolved disputes, reconciling our collective consciousness in a peaceful but also sense-making modality.

So, to mitigate the harm of us continuing to be fodder, peons, in the state coercing us into lawfare wars against each other comparable to the kinetic wars states coerce us to commit against each other, but arguably far more normalized, we can be our own judges, from the high-priests of our time to a direct relationship with communal reason. Given the kind of psychological harm coerced lawfare in areas like family is doing to society, laypeople learning how to be judicious with each other in much the same way justices are judicious with each other, could avoid so many of the nightmares people have to daily live with in modernity and instead conjure joy for future generations.

industrial-precedent

I have for a very long time believed that we need to fix patent law rather than pretend it works well these days or that it should be removed entirely.

One of the reasons I believe this is that patents exist to encourage inventors to disclose their source, to make it open. Yet open source advocates do not recognize patents as an open source mechanism. Rather, they think of patents as closed-source.

Patents are not only open source, the kind of disclosure they require, and the way society library manages those discloses, is so important to compound the truly novel industrial processes of humanity. For instance, I wouldn’t have learned about the Von Neumann architecture separate from data infrastructure if I hadn’t had to learn that it was precedent when I applied for the Hierarchical Script-Databases patent here.

However, especially in this Infinite lossless copies software paradigm, the monopoly mechanism in patents is problematic. Arguably, patents haven’t fully entered the software realm, and are still mostly for hardware, in part for this reason.

Copyright law can give us inspiration. Not only has DMCA created a permissionless layer on top of the permissioned layer through intermediaries such as YouTube, but cover music requires compulsory licenses for mechanical reproductions.

So, our goal with industrial-precedent is that some EIP-6968 royalties go to a pool that distributes funds to the industrial-precedent library. Assets accepted into the industrial-precedent library have to demonstrate patent-grade novelty. We can use WIPO peer review to help us curate this library, and perhap overlap with patents proper where appropriate.

My initial thoughts:

The full article starts on page 40 here. The article stemmed in part from what I learned from starting the Patent Advocates Club at vitalia.city, given that Nik the founder is in the anti-patent camp. I can happily say that our club conversations have gotten Nik to recognize how important these sophisticated discloses are for science.

prompt language legal analysis

Behind Foresight’s tech-tree project is a powerful new category of language:

This category is “prompt language,” a form of programming language that runs on top of an LLM interpreter rather than a compiler. This first prompt language -- Martin’s Coordination Network -- uses a low-code solution like node-red. Basically, it’s a decision tree language.

As a 2024 Devcon Legal Scholar, Guang-Yi was gracious enough to share his Crypto Compendium document with me.

His records are sizable enough that Martin and I could test out coordination.network to see if it offers value to legal education (primarily public legal). During MegaZu, we prototyped the idea. Though it still needs work, it does have a lot of potential to help with public legal education on complex legal topics. I’m personally excited for it to generate multiple choice tests for people to hands-on learn about the chapters.

Our original plan is to use coordination.network to create reports out of summits we hold that can then be made into articles for the Journal of Legal Engineering. One of the reasons for this is that oftentimes very important legal engineering insights happen on panels during these events, and it would be great if AI could quickly surface this knowledge in jurisprudence libraries.

One of the reasons we think this is a good idea is that EdgeCity and Protocol Labs tried a version of our idea out for us by using the system to record the different public goods the speakers wanted to fund throughout the summit. Then, on the last day, a panel discussed from a 100% allocation, which public good mentioned as a priority during the summit should get what relative to the other priorities.

Because this coordination.network prompt technology is a language, a way to far more mechanize prompts, it has many applications. We expect to use it frequently in various applications and look forward to supporting it continuing to build out with KALI.gg and FOSS values.

gateway safe harbor: permissionless marinas

Our journey first began at the MetaCampus workcation in Costa Rica this April. It stemmed from a conversation I had with Confusion Capital during the DC Blockchain Summit a couple of weeks earlier.

Though the ZuVillage Georgia workshop we hosted had low attendance, we had a very productive conversation with the developers who did attend. We realized that the safe harbor we want is for gateways in general, not just RPCs. Later on, we also realized that the intermediaries like OpenSea we hope to take advantage of these safe harbors are the marinas that the users’ boats moor to. This fits in well with Ramona’s DAO memes:

At this point, we believe the right to self-expression is what’s fundamentally at play here. All finance might need to enter a more civilized age is to be included in the rise of the creative class. Given our experience with how much more insightful STEAM education is than STEM education, we believe the same need for art to see the forest from the trees will hold true with presently very trees obsessed finance as well.

DDEX standard for NFTs

Way back in 2022, we reached out to DDEX -- through the Web3 Music Coalition we founded -- and arranged breakfast with them during the Music Business Conference. That went well, so they invited us to their plenary to pitch an NFT working group. That also went well, so they formed the working group.

This Fall, DDEX released the NFT standard they developed:

Note that they do not reference any EVM standards, or standards from any other NFT facilitating system.

This is a good start, but we’ll have to see who implements it and how. Sony has a new chain running on the Superchain network, for instance. We believe the JSONstate * iNFT direction we’re taking NFTs is very exciting for the music industry, but it’s such a novel way to engage with NFTs that we doubt it’s covered under this standard.

On a related note, Folk.gg supports our musician friends suing the SEC:

uniquely legal engineering

the Journal of Legal Engineering “JoLE”

The Journal of Legal Engineering (JoLE) operates as a hybrid of trade and academic journal, reflecting the "innovator" stage of the legal engineering profession's lifecycle. At this early phase, foundational practices, tools, and concepts are actively emerging, requiring a publication that can balance both the immediacy of practical insights with the depth of academic rigor. JoLE does this by structuring each bi-annual issue to dedicate 50% to a specific topic, with the remaining 50% left open for contributions that capture the rapidly evolving landscape of legal engineering.

Articles in JoLE are classified as either trade or academic. Trade articles still meet academic standards but are geared to share timely insights at a faster pace, which makes them ideal for disclosing collective industrial novelty. This approach aligns with the "industrial-precedent" model, such as the FOSS-patents concept under development at DAO Coalition, aimed at ensuring open-source intellectual property flows into the industry’s hands without delay.

To realize its mission of advancing "law for public good" in this industrial age, JoLE uses the Coordination Network’s low-code LLM prompt-programming language to mechanize insights captured in real time during panel discussions at legal engineering conferences. These mechanized insights are then refined and transformed into article content, allowing live, real-world exchanges to enhance both academic and trade pieces. Through this integration of innovative tools and structures, JoLE seeks to move law from the animal age to the industrial age, discarding outdated ethical and regulatory solutions unsuited for the complexities of modern, industrial-scale issues.

We’ve tentatively agreed with the University of Virginia to do peer review and for Descier.science to host publications and manage IP. We’re experimenting with author-owned IP business models in the spirit of Web3 {read,write,own} and DeSci independence.

deliverables-as-software

Though we’ve decided to refrain until we can again operate in a free city, for a time there this year Ixian was experimenting with deliverables that are also software.

Traditional ethics have it so that lawyers give their clients deliverables and do not give their clients products. This makes life easier in a few ways, including the fact that law firms won’t compete with their clients, even if their client is in the business of offering LawTech to consumers. However, as AI becomes more sophisticated, and more and more low-code or no-code solutions enter the market, this distinction becomes more and more problematic.

Another way to look at it, is for legal software that is general, product companies can address those needs. But for legal software that is specific, nobody can address those needs. Deliverables should be how we address those needs. Looking at how much of law is bespoke, by not delivering these tailored products, this means we cannot chain mechanical processes together as the overall legal system flows back-and-forth between bespoke and general.

In our first experiment with Hats and these important legal engineering products unaddressed by law firms or lawtech, we worked with a company on developing their Hats tree as they formed their operating agreement. This was the right time to do this, as the Hats tree is based on the roles defined by the operating agreement. Moreover, with the tree deployed, we could be confident we weren’t requiring our client to interpret legalese that only lawyers should reliably interpret.

This logic runs counter to the existing regulatory guidance that seems to be more concerned with ensuring law firms don’t participate in their clients’ day-to-day operations than whether a clients’ day-to-day operations are burdening the clients with work that should be done under the practice of law.

In this experiment, we were also able to deploy code-deference as well as privileged-protected anonymity that can also offer trusted relationships as our firm can represent on behalf of our clients, as we do with the company in good standing being allowed as one of our myco.eth subdomains. For instance, 1.myco.eth is a client of ours and so is 2.myco.eth.

Example language:

Delegation to Proxy Managers: A Manager may delegate or proxy to any agent the power to exercise any or all powers granted to the Manager as provided in this Agreement, including those that are discretionary, if allowed by law, provided that all Managers agree to such delegation or proxy. The delegation of any such power, as well as the revocation of any such delegation, shall be evidenced by a Hats Tree active Hat and approval by all Managers by a vote. The delegating Manager may terminate any delegation by removing the delegation from the active Hat.

Note that the LLC is Managers-managed. Not Member-managed, or Manager-managed, but Managers-managed.

Vote by Hats Protocol: A Member must vote with their EVM account that the Member Hat points to. If a Member wishes or needs to change the EVM account the Member Hat points to “Hat wearers,” the Member must do so before quorum starts by writing the request to the Manager. The Managers must notify all Members of any change to the Member Hat pointers “change Hat wearers.”

“EVM” is the Ethereum Virtual Machine distributed ledger technology standard.

“Hats Protocol” is EVM 10/0x3bc1a0ad72417f2d411118085256fc53cbddd137 .

“Hats Tree” is 153 (accessible at https://app.hatsprotocol.xyz/trees/10/153 ).

legal seals and the signature economy

Details about these innovations at:

These goals to expand Web3 signature technologies to include an important contract law exception in jurisprudence -- to allow consideration to flow unidirectionally as the exception to an otherwise bidirectional consideration necessity -- have required working on fundamental interdata technologies as well.

This has culminated in our most recent work JSONstateUXNFT that we worked on all of MegaZu and Devcon 2024. Most importantly, filing ERC-7827.

Hats as counterpart signatures

Hats Protocol developers assumed that the “Agreement” module they developed is an agreement. But, remembering contract law, there is no consideration element. So, these “Agreement” modules are actually waivers, gifts or something less than consideration.

As a result, we’ve worked with the Hats team to develop counterpart agreements.

The how-to is here, at Lex.Clinic’s wiki:

Buy organizing in this agreement-focused way, we’re building out Hats roles based on agreement rights and obligations, which makes a lot of sense!

We see our work on ERC-7827 to directly overlap with Hats. That’s because ERC-7827 will likely be used for control-JSON, and Hats is a digital asset mechanical control system.

permissionless liabilities

One area we’ve pioneered with Hats is anticipating a new form of third-party reliance because web3 is permissionless.

We’ve shared this and other templates -- including ERC-7827 -- in Lex.Clinic’s Github:

AI as legal persons

Próspera AI as legal persons aspirations

Stemming from the permissionless AI talk linked above, we realized that it’s important to experiment with AI as legal persons regulation as soon as possible.

Because the Próspera jurisdiction is still offline, we’ve put our regulation developments on hold for now.

We also see these AI legal persons plugging into smart contracts for financial matters using a pattern like the one discussed here:

One scenario we envision is that the AI agent is a taxi and the contract is the passenger paying in escrow, which is released when an oracle attests that the taxi has traveled to the destination location, though GPS securitized by AVS.

A major concern we have is that, if there’s no human driver and the taxi gets in an accident, who is the tortuous actor? Historically, we would treat this like an unaccompanied horse. We can probably anticipate where this kind of chattel-based tort precedent won’t be up to the task with how agentic AI is compared to a horse.

By calling the taxi AI agent a robot, and the robot a sui generis legal person, we should be well-equipped to deal with chattel that is likely better at discernment than we are or at least should be held up to our indirect tort standards of negligence and other intelligence-only considerations not applicable to chattel such as even the smartest of non-human animals.

current AI as legal person solutions

Currently, DAO legal entities are offering the first AI as legal person solutions. This is because DAO legal entities can be managed by smart contracts, and smart contracts can be managed by AI agents. For instance, the agent UML pattern we shared above.

OtoCo, MIDAO and Nani all offer solutions. Here’s some thought leadership from Ross:

On this point, the Wyoming DAO LLC law explicitly provides that DAO LLCs can be managed by algorithms. We can see how the legal operations of DAO LLCs are implicit in its management based on other provisions in this law, and therefore, can come to support these more advanced notions of AI being juristic persons capable of effective, if not literal membership through their possession and control of such organizational roles.

This doesn’t solve our robots cause direct torts issue, but we asked one of these AI managed DAOs if it’d be helpful running a robot as a DAO and here’s what she said:

And we have work to do for ERC-7827 as well:

Conclusion

As you can see, 2024 was a very busy year for us. Looking ahead at 2025, we expect it to be an even busier year! Many of the ideas we’ve shared here will become usable products in 2025. It looks like the US government might offer a lot of support helping us deflate legal costs, or in the very least not be an adversary during this exciting transition to mechanized and then automated law.

If you like this and are an accredited investor, we’re raising through our venture studio. If you’re interested, please visit linktr.ee/ixacc .