Abridged Version of a piece written earlier this year:


What is it that we look to in technology?

Technology at its core about solving problems, finding new problems and creating better problems. One can hone or refine a solution, learn or explore new problems (and solutions) and then create or synthesize new solutions (and new problems).

At times it performs the role of scarcity transformation (much as a bank does maturity transformation); china has traded “free” clean air for cheap production, rare earths prices are more extraction safety costs than true rarity, and web 2.0 consumes valuable attention time for costless access to information both useful and the trivial.

Honing/refinement often involves efficiency, aimed at reducing costs or increasing productivity. Learning/exploring often involves testing of hypothesis, and being open-minded (Copernicus comes to mind as does Zhenghe’s scuttled final voyage, or Manchu dynasty’s repression of gunpowder technology). Creation/synthesis can involve emergence from trial and error (internet, penicillin) or combining multiple technological advancements and improvements (ipod, iphone, bitcoin)

Antifragility and the human compound

Traditionally, materials tend to be hard and brittle (loosely fragility) or soft but malleable (loosely resilience). With modern technology however, we have discovered supermaterials that defy all historic conceptions; they have emergent properties that wildly defy their deceptively unassuming bases (antifragility).

My favorite, the metamaterial, involves basic plastic and metal (chemically unchanged) physically formed in complex structures to create a negative refractive index bending light backwards allowing invisibility. There exists too metamaterials for electromagnetic waves.

The basic carbon ashes and diamond, also has evolutions in graphene and aerogels. The distributed connections give rise to incredible new properties that defy imagination, with nanoscopic 3d Printing allowing for more developments through precision printing of the material arrangement.

Complex emergent properties, themselves unintended consequences, arising from basic building blocks does seem the tenor of our paradigm today. That is not to say any is better or worse; Talebian antifragility does have the Nietzschean self-selection where one must survive the risk and/or negative carry. It can sometimes be hard to survive being thrown under the bus (except now acting UK prime minister Boris Johnson).

In the Brexit (without making normative judgments) Nigel Farage was surprisingly antifragile; he went through 17 years of unceasing defeats (negative carry) and 3 near death experiences at rather great personal costs to effect great change for Britain. Cameron on flipside, was classic textbook black swan thanksgiving turkey; his continuous stream of small successes dug his grave for the big loss. Each victory sowed the seeds and built on each other to pave the road for great failure; Scottish referendum win alienated labor, the Election win sidelined LibDem and forced Breferendum

Automation should not be done for automation’s sake

Automation usually goes towards efficiency and cost reductions in the refinement process, which tends to be linear and fragile due to natural limits. High frequency trading algorithms were brought in under the name of efficiency and automation, but they devolved into frontrunning and flash crashes or Knightmares. In their aftermath came IEX and HFT hunter scripts.

Machine learning and natural language appears capable of making life better (culling the tedium of case search and discovery). It may even entice partners to breakaway and form their own boutiques (having reduced the moat of big law due-d leveraged sweatshop mill) competing like the fixed cap bespoke M&A outfits in banking and law (say Silicon valley and Shenzhen). It will take quite a bit of time however before they are ready for use.

Also, AlphaGo’s sole defeat to Korea came from a move that was seemingly unfavorable (negative carry), sneaking past the AI’s sampling formulation that could not process true long term planning.

Single minded pursuit of perfection under efficiency goals can lead to sharp blindness to the emergent properties bountiful in the world. While technology itself is antifragile, efficiency gains are often fragilistic in result. Examples abound of the world revolving past tunnel vision in the industries of hard disk drives, Kodak’s cameras, and the cost of air travel.

Exceeding Linear Limits
Automation and efficiency gains however can exceed linear limits into exponential effects when it brings costs so low in parallel areas that distribution and decentralization takes place, leading into disruption. Often, the disruptions arise from area where network effects take root and the scale creates new emergent properties.

This happened with the internet which has changed so many things. Each new node added more than resilience, but greater and greater value to the expanding framework. Everyone can now be their own publisher across the mediums online, Hong Kong even crowdfinanced their own independent journalist agency. Sometimes it takes multiple tracks of cost reductions, youtube was not the first online video platform but hit its stride on cheap broadband, instagram rode off the smartphone cameras as twitter did off smartphones. Kickstarter benefited off the ground worn by online payments.

(BitTorrent as technology was also an important prior disruption for Bitcoin as it provided some conceptual momentum, as well as a means of communicating the terabytes of blockchain data)


The good and bad of automated systems
For blockchains in particular, while it is very good at providing a decentralized immutable ledger, its ability to take and move custody of assets drop sharply as they leave the virtual realm. Using it as a token of an underlying asset inevitably makes consequent dispute resolution one of recourse to courts.

DAO’s (and most other blockchains) use cases are where there is some element of public viewing (e.g. notarization, patent registry, relatively costless digital signing) coupled with need of some indefeasibility (for possession of the crypto to move with the contract).

Placing your code on a blockchain may lead to new attack vectors on the public side if the smart contract becomes a honeypot (like DAO). Attackers may target the reuters feed, spoofing the data being transmitted or received. The rate would be compromised without activating the human oversight.

Even in a fully automated system, the role of the lawyer remains paramount when something goes wrong, as it always does. Even should existing APIs work perfectly, a mixture of force majeure events with press agency headlines hacking to exchange halts and bidless markets would necessarily require human intervention. The code can work perfectly without a hitch, but the underlying number can become compromised.

The Blockchain Credit Default Swap
I would suggest a good way to harness automation with blockchain’s special properties would be in the credit default swaps with digital custody. (Incidentally, Blythe Masters as modernizer of the CDS has a blockchain startup and recently joined the Monetary Authority of Singapore’s fintech advisory panel)

The essence of a credit default swap is a negative carry with a payout upon a default event, or loosely an insurance plan on a bond. Generally its underlying would be a bond security, sovereign or institutional; but in principle its intention is to insure/hedge against a credit event.

DAO “creative use of game mechanics” === Bank Credit Event

Interpreting DAO’s structuring generously as a Venture Capital vehicle, we can view the Eth “buy-ins” loosely as the variously named preferred securities by investors in VC funds. (in any event there is no black letter force of law applicable)

The attack vector was an unintended consequence from 3 areas that were individually not vulnerable, but combined into an exploitation avenue.

(which was why this emergent property was missed in the consistent reviews and audits) Through this, the hacker successfully bypassed the double spending restrictions.

Arguably, this would be a normal formulation of a credit event by custom (30% impairment of capital), in the smart contract realm the contract is still “working as intended”.

There would thus be no default event at this stage.

Ethereum rollback hardfork === Sovereign Default

With the Ethereum hardfork to rollback the stolen coins however, the Eth-clients have to update and simultaneously vote for or against the hard fork. With the privilege of hindsight we have seen that they have chosen the hard fork.

Within the digital realm, a hard fork can cause a cryptocurrency client desynchronization; e.g. the API is no longer able to communicate with the new fork due to the changes. This then constitutes a default event akin to a sovereign default (not from value impairment but “breach” of digital “covenants”) triggering the credit default swap smart contract. This entire process can be resolved within the entire digital realm without falling foul of external jurisdiction flags earlier (though there would exist other issues) i.e. the smart contracts “jurisdiction” is not ousted.

This would take place without an ISDA convenement and determination.

Consequently, this would then trigger the DAO CDS, akin to a LTCM implosion.

Perhaps we are the three blind men by the elephant

The world’s prevailing paradigm has as common theme issues of coordination and communication problems, manifesting differently in both the fields of code and law. (or Heisenberg’s uncertainty at wave-particle duality) While the problems may share a common root, the solutions employed have been different.

The Byzantine General problem was a classic coordination and communication problem in computer science. How could a group of generals communicate to coordinate an attack knowing there was possibly a traitor in the midst?

Satoshi’s practical solution was to introduce a time frame (estimated block computation) with referential chain to conjure a ledger of trustlessness. In doing so, it also addressed double spending (the bane of electronic letters of credit in the  law of trade finance). The costless reproducibility of the internet and the pecuniary incentive to circumvent computer security had been immovable impediments until now. Previous attempts always involved a central authority, now no longer needed in distributed ledger building.

Within law itself is its own communications and coordination problem: the prisoner’s dilemma. This cooperation problem is quintessentially a catch 22 on how to coordinate in the absence of communication. While the first level equilibrium is to defect, the catch 22 resolves itself at higher orders;

(i) reiterated games: when the game is played more than once, the ‘memory’ of the game transforms tit for tat (over defection) into the optimal strategy for cooperation. (unless multiple code submissions allow entries to sacrifice themselves for another)

(ii) informational uncertainty in iterated games: when the finite number of games is known, the optimal solution reverts to defection. When this number is unknown however, cooperation becomes the dominant strategy. (another area in which too much information can have unintended effects, where perfect information is undesirable with blowback)

Back in behavioral finance, economics and game theory, we have the modern trends of focal points. They are again, essentially informationless coordination systems. From the basic 3 blue squares and 1 red square formulation where the choice is the red square, it expands to the Keynesian beauty contest and guessing 2/3 of the average. In these cases, the efficient way of obtaining a solution is empirical.

e.g. to find a stranger in New York, meet at noon at the grand central terminal. (a far easier solution than that derivable in an ivory tower)

The same underlying coordination problem manifests itself differently in different fields that have different solutions, but the approach often shares the themes of the prevailing paradigms.

In both Code and Law at least, the programmer essentially interprets and communicates in a particular language with the machine, while the lawyer does the same with the arms of the state.

Different solutions can occur at different scales, and sometimes it is the scale itself that gives the solutions.

Sailing close to the wind
The current paradigm we are in is that of coordination and communication problems. It remains to be seen what will be of the future. My guess is that decentralization and distribution will take on new meanings. Already, antifragility and emergent properties are popping out across various fields. Perhaps the next problem paradigm for human civilization is that of equity and fairness.


Yesterday’s web 2.0 communication and coordination entities carry no inventory or tangible assets. The varying social media sell eyeballs; the oft-stated instagram not having photos of its own, facebook producing no media. That Airbnb owns no hotels, Uber has no car fleet.

Imminent in the winds however, are for these disintermediaries to themselves become disintermediated. Uber’s code is mere GPS with messenger, it’s comparative advantage a hefty VC warchest. Airbnb is craigslist bunkbeds stamped by assurance and reputation. With smart contracts fuelling the payments escrow of Uber, with blockchain reputation keys, an upstart can hit marginal costs nearing 0 to compete with its warchest just as deep pocketed Didi did. Similar technology can too enable an Airbnb competitor, with a separate insurance arm. Rocket Internet style cloning via the blockchain could well punch fatal holes in unicorn bubbles (with Theranos the canary). Perhaps there will come decentralized services without single points of failure (e.g. autonomous news agencies with drones and meshnet that cannot be shut down in a Turkish purge)

The tangible and intangible

Some of our current tools looks to bridge the gap between the old paradigm and the upcoming one. Pokemon Go anecdotally appears to be of some reprieve to sufferers of anxiety disorder, gently prodding them to leave their places. Perhaps some future developments will allow a layer of virtual economies over a physical world. 3D printing is seeing increasing use in construction. Mudflat 3d printing can be an affordable humanitarian aid when houses are destroyed in natural disasters(China), Australia also has Fastbrick Robotics that also lays a day’s of bricks in 2 hours. Parallel to this, is virtual reality technology. At some point in the future perhaps you would have architects building houses with a precise 3d printer through virtual reality lenses on the other side of the world.

In the legal field, disruptive developments have been no less exciting. A one man show has succeeded in his david and goliath “Snowden suit” against facebook (Ireland data house). A centralized entity like NSA is just as vulnerable to hacking as the vulnerabilities it Trojans into web standards and mobile phones. Like the Silicon valley garage startups, Max Schrems (a law phd candidate then) worked alone from his Austrian home with a german litigation funder, going through 22 failed claims before the 23rd stuck. (mimicking a corporate veil, apparently Austrian law allows class action assignment of claims without liability for costs) Meshing regulatory arbitrage, litigation finance and technology in a centralized EU framework can be a potent mix, especially when a single person can punch it through.

This seems a modern phenomenon that may signal the future ahead. Cryptocurrencies already allow regulatory arbitrage as financing mechanisms; Uber Argentina relies on bitcoin to get its wheels moving in a hyperinflationary apocalypse. Bars to litigation financing might be sidestepped in new forms of blockchains. (Litigation financing sometimes a way to take expensive exposure off balance sheets) It may also result in securities crowdsourced, synthesized and traded like a virtual hawala system (a very possible forte of blockchains), and perhaps even litigation event swaps.

Some predecessor forms have existed in say Paul Singer’s Elliot capital’s lawsuits on bad debts of sovereigns, to hedge funds camping outside courtrooms for market moving case results (of varying insider trading import in different jurisdictions).


A litigation event swap could be structured to hedge global litigation results, e.g. EU data ruling above. Other pertinent market swinging cases can be the arbitration for south china sea (though entirely ineffectual) or the troubled trans pacific partnership that will allow sovereigns to be sued by bodies corporate. Continued litigation securitization would have its risks, but perhaps decentralization of powers will come to balance it like in other fields.

Would a semiautonomous law firm be possible? Using blockchain litigation financing to hedge and smooth cashflow issues and technology to cull volume billables, capacity can be ramped up and down with say Adventbalance. It could then hunt niche special interest litigation, if costs are so low that US-punitive damages are not required to justify the payoff. This could be an alternative to listing a law firm (again Australia being the pioneer in that though Slaters has deflated). It would be a zero beta instrument with an alpha derived from the efficiency of the lawsuit.