Internet of Agreements 2 - Digital Catapult

Andy Park:

The hashtag for today's Internet of Agreements conference is #IofA - schedule is here

Robert Learney:

I'm Rob, and I'm the lead technologist for Blockchain at Digital Catapult - we've been around for about 5 years; we're here to encourage digitisation of the UK economy.

We work on lots of new technology - VR and media capture, 5G and AI. We're just getting started on Blockchain here.

Vinay Gupta:

Our topic is Blockchain and World Trade - these have a lot in common. Keeping track of everything in trade is a huge problem, and is still mostly done through paper documents

Anyone who has done anything with an invoice finds that everyone involved has a differerent number for the invoice

Blockchain has the potential to keep track of all this information in an accountable and public way, that can combine payment with a register of ownership

We're moving into a new economic phase with more focus on regionalism than globalisation - being able to verify compliance with regulations automatically could be Globalisation 2.0

Robert Learney:

At Catapult we have recently launched the Machine Intelligence Garage too

Vinay Gupta:

if you're here, do a tweet with #IofA to say what you are interested in

Stephen Palley:

I have been a lawyer since before Satoshi wrote his paper - sometimes it feels like being in a Dickens novel; I built a dispute resolution system using rails and postgres, and realised the trust issue

I have clients who are trying to build blockchain technology that si legally compliant - I like Vinay's "you do you wherever you are and software will figure out compliance"

Clive Freedman:

I am a barrister and have seen more issues resolving disputes over blockchain

Adam Sanitt:

I'm a lawyer at Norton Rose Fulbright and doing work with ledgers there; I also research and public in mathematics, medical statistics and open source code

Chris Wray:

I'm chief legal officer at Mattereum.

Trade is the negotiation of reciprocal promises that are enforceable - that is a legal definition of contract, but it is what trade is at heart

Negotiation was originally in person or local; technology make it work over longer distances.

Reciprocal promise means that I will do somethign for you, if you do something for me.

Enforceable means that if you do not fulfil your promise, you are required to compensate me.

I'd like to put blockchain technology in this historical context of contract agreement. IT has helped automate the negotation of promises, and the settlement, but maybe less impact on enforcement.

So what is new about Blockchain? How does it change what can co right and wrong with trade?

Clive Freedman:

Putting your contracts on the blockchain would mean that software could act on a trigger for you.

The difficulty is that a judge can't read the software and decide what you have agreed. Commenting the code may not be enough to clarify this.

a more useful way would be for an expert to explain the details of the smart contract. What may make more sense is a parallel human readable contract to accompany the smart one

this will enable the setting of arbitration terms, venu and country of law

fact A causing action B to occur can go wrong in several ways - fact A may not be detected; action B may not occur because the computer may be offline

Now many of these things can happen without blockchains or smart contracts - look at margin calls for example

or look at investors being able to buy invoices from retailers to spread their rick

there is likely to be a 20-30 page contarct drawn up by lawyers to specify how disputes are resolved in these situations

If we're over-reliant on the smart contracts without the accompanying natural language contracts in place, thinsg may go wrong

Stephen Palley:

Everything has changed and nothing has changed. Look at the example of a Letter of Credit. - what if the bank behind the letter of credit is no longer solvent?

The notion of guaranteed enforceability is attractive - you can get a judgement without being able to collect it; maybe blockchain can change this

Contract law itself was an important technological innovation - changing from self help to state power. Can blockchain take enforcement away from state power?

I was asked recently if Insurance covers a deal made on a blockchain - I had to explain fork risk

how much was frozen in parity? $100M ? Its only a matter of time before sums like that end up in front of a judge?

Adam Sanitt:

The problem with international trade is that you are separated by space and time. We try to replace physical exchange with letter of credit or bills of lading

Blockchain lets you do this exchange digitally, and keeps a record of the transaaction

I think of reality mismatch errors - the blockchain is correct in what it records but reality is at fault. We can't avoid reality mismatches

another reality mismatch can be a transfer made in error or with a gun to your head. The courts will reverse these errors

when contracts were written down people thought we would no longer need lawyers to handle them. The same will happen with smart contarcts

Chris Wray:

there is an idea of Ricardian contracts - a piece of software that enforces a contract but comes with a written legal agreement that covers it too

if we have contarcts that are perfromed automatically, they may no longer need enfircement

Stephen Palley:

a couple of years ago you would see that "code eats law" or "code is law" - contracts are an artefact of trade.

if you could create system that removed any possibility of a a disagreement that may be a problem too. Sometimes you want ambiguity in an agreement.

there is a difference between an agreement and a contract and a dispute - contracts assume enforcement is there.

The point of courts is to have a neutral arbiter deicide how to resolve the agreement

Chris Wray:

we can have contracts without a written record agreed, but it may be unwise when convincing the state power to enforce them

if our automation gets good enough would eb be willing to do away witht he parallel written record?

Tju Liang:

we may move away form the environment of a contract that goes to a court is move it to another enforcement model

Right now we enforce IP for example in a joint venture in hong Kong/Singapore -

I would say that law is code rather than code is law - to properly draft a contract you have to have a flow in mind of the transactions going to take place

Tju Liang:

so you can get a programmer to implement the legal agreement, and that will test the contarct too

Chris Wray:

Lawyers didn't have the opportunity to define business arrangement so that a dispute was less likely to arise in the first place

Stephen Palley:

Ultimately lawyers are going to learn how to program. Going to court is fun, juries are fun, but we will see less of them

Tju Liang:

contract drafting is a game now where people bugfix contracts to cover new situations over time

Chris Wray:

we've had the 'code is law' - code is just code, but law is system design - there are different scales or levels to carry out those tasks

Adam Sanitt:

we have moved from the future of law to the future of lawyers

we have found that in many cases the cost of legal recourse is too high compared to the initial transaction


the other key aspect any of this working is consent. Both parties need to consent to the terms too.

humans have 2 mechanisms of enforcement - violence and discretion. Courts employ violence; rating systems like ebay ebody discretion

Chris Wray:

we need to distinguish between the retail level and b2B. We lose the possibility of escrow in international trade

Tju Liang:

I'm working with a team trying to build a blockchain exchange system for Africa

I'm not sure that reputation scores on ebay are any different from credit ratings

Stephen Palley:

Trustless isn't a good term - we're moving the trust to software ratehr than individuals

Tju Liang:

the purpose of trustless is to move trust from those who are party to someone neutral

Stephen Palley:

but you're not moving trust to neutral people but to those who have a stake in a different way- miners and coders too

Adam Sanitt:

you're moving the trust to a collective endeavour - it depends on which blockchain your on. You don't need blockchain to have a reputation system

the analog within the blockchain system is to exclude bad players without recourse to the legal system - it gets more difficult in multilateral situations


I'm an international oil and gas lawyer - I can see how it works for a single transaction, but how could this work with IP rights?

Chris Wray:

there will be things that we can fully automate - if we'r exchanging bitcoin for ether, that can happen instantly; if it is oil I'm supposed to receive that is different

Stephen Palley:

a quantity takeoff in construction is the dirt you need to remove -this is debated over volume versus weight. Using a drone and photogrammetry you can calculate this more accurately

Take the most complex process you have and break it into smaller pieces and work on those with programmers

Chris Wray:

what does an economy look like when nobody can get consequential losses?

Vinay Gupta:

we get into a fuzzy middle ground where reputation, insurance and escrow end up interacting with each other

I tend to think reputation systems are very fragile, but I prefer to look at insurance and escrow - how do we move from an escrow norm to an insurance norm

Stephen Palley:

escrow and insurance are different in the us - escrow is a 2 party method; insurance has a 3rd party involved.

Tju Liang:

from the buyers standpoint they serve a similar function - I want to get paid

Tju Liang:

you will see, in M&A contracts for example you will see insurance for breach of warranties


I'm not a lawyer, but I get upset when people talk about trustless without defining what it means

as an ex banker when I look at trustless agreements they look like fully collatoralised agreements - banks came into being to avoid those

Adam Sanitt:

We brought up trustless initially, but it's not anarchy - there is trust in the fiat currency and in the other party - you can't see where the trust has moved too.


I have a slightly challenging view - as a lawyer you learn that contracts are about relationships, and most of them aren't enforced

the systems that are most useful are anglo-saxon legal systems not civil systems - there are equitable remedies like estoppel if you don't perform

I'm not convinced that using blockchain technology on its own - you want to encourage trust not to trade trustlessly

Chris Wray:

Trust but verify might be the right way to think about it. If you have evidence you can take those and hash them and put them on the blockchain too - you can resile from them then


the world changes when people decide to do things in different ways - they negotiate partial completion

Tju Liang:

trust does not fully scale - when you buy a random object from ebay you are trusting ebay, not the random seller

Stephen Palley:

if you want to maintain a mercantile relationship, discretion is key

Adam Sanitt:

not only are most contracts not enforced, most of them are not even read until a dispute occurs

we're not advocating a lack of trust, we're just adding a different trust arena


I was an american international lawyer - everything is in the context of the UK and other countries - in the USA its different. International letters of credit were torn apart by the 50 separate states int he USA

Stephen Palley:

The US is not unique in having a federal system, but contract law is different in the different states - insurance policies have to be written in each separate state

unless congress takes something away from the states, it is not separately enforced

Chris Wray:

with all this fragmentation, don't forget the wonders of arbitration, where you can choose a venue and expertise

Rob Knight:

Before we used to keep track of gold in a vault, and assign ownership of it through a ledger

with bitcoin, possession of the key is possession of the asset - what we're trying to do at mattereum is apply that representation to physical assets

we want to combine a legal agreement with a technical system that keeps track of the assets

we have moved from physical systems of money to payment by credit card and electronic money

with credit cards there is a way to refund on non-delivery, and a form of identity available for online commerce

what if we brought some of these extensions to blockchain transactions - could we bring dispute resolution and identity to company to contracts

could we make it as easy to sign a mortgage or a rental agreement as it is to use a credit card?

when we talk about blockchains and world trade, we are talking about making it as easy as a consumer purchase using a credit card

Mihai Cimpoesu:

I'm Mihai Cimpoesu, chief engineer at Mattereum

Vinay Gupta:

there has been 20 years of work getting computer properly integrated into trade and working reliably on the internet, but we still cant get a net 90 invoice and payment automation to pay it

Chris Wray:

it feels like there is scope here for lawyers to do more - to do coding in effect, and to do systems engineering and scenario planning too

It would be preferable for all the legal work to be done upfront, so the software can become standardised

pushing contracts to be more standardised is a constraint on freedom of contract. That standardisation requires upfront a systems design to reduce the number of dispures that arise

if you get this standardisation right you can anticipate the things that could go wrong and fix them over time.

I think there will be whole classes of smart contracts that will be written as stable ideas that are as natural as using a vending machine

Vinay Gupta:

Tell us about the technical side of this, and how coders analyse contracts

Mihai Cimpoesu:

It has been a significant mind shift for me from turning vague requirements from a client into code, to coming up with a broader design of a system


can you expand a bit on how lawyers who are late adopters of technology can adapt to this?

You said the requirements process was different - can you expand on that?

Here there is a deep cultural background of lawyers based on their century of training and history - how does that affect it

Chris Wray:

when you are dealing with multiple jurisdictions, even though you can choose venue for arbitration, there will always be a need forlegal thinking and invention

we do development of law and of code in house, but we still need input from different jurisdictions

Vinay Gupta:

often a contract is triggered by a payment arriving or time elapsing, which should be simple enough

but in practice the complexity of the real world means that we are connecting a paper contract with a smart contract

Mihai Cimpoesu:

clients would come with a half page sketch and say they wanted a blockchain thing; i would turn it into 5 page technical spec

Chris Wray:

which would become a 10 page legal spec

Mihai Cimpoesu:

we would then write the code and the contract and go line by lne through both computer code and legal code.


how reliable are the underlying parts fo the infrastructure? If you think of DHL or Maerske, how long before they will be on these?

Chris Wray:

aIe don't think we can get complete interpretability between written language and computer code

Rob Knight:

you can do things now on private ledgers easily enough; doing it on public ones is much harder

Web Development brought a huge number of people into software engineering, but not always wiht great ethics

the Facebook 'Move Fast and Break Things' approach does not work with smart contracts. you cant ship it in the morning and fix it in th eafternoon

with git blame there is a level of accountability in smart contracts that there isn't in web dev

Vinay Gupta:

millions of dollars have gone astray because of Ethereum's engineering practices

I would expect to see software reliability techniques from aerospace move inot blockchain and ethereum

crypto currency is like saying horseless carriage - there was a period where knowing what the web was and explaining it was a job - we're still there with blockchain

blockchain is a process, nto a concrete thing - it's like saying 'sport' not cricket

right now in terms of getting large scale systems to run they will be the most conservative systems that take off

there will be an equilibrium between the legacy people who add a thin layer of blockchain, and the blockchain people gradually understanding legacy process

Google, Amazon, Microsoft, SAP et al have been very quiet about Blockchain

Rob Knight:

a lot of my thinking about smart contracts came from digitising BBC contractual rights and working out the implications of 1970s contracts for streaming and download

those challenges are often more difficult than the engineering

when I saw smart contracts I thought wouldn't this be great for doing settlements of rights and payment for those kinds of royalties

If you are a large media organisation, there probably is a blockchain smart contract future for your organisation

there are a lot of proof of concept and pilot projects going on - companies that have adapted to the web will adapt to the blockchain too


There is some interest in government in looking at these technologies, but I expect the private sector to move faster

there is interest in the DoJ using Blockchain for storing the chain of evidence for exhibits with hashes on the blockchain

this used to be done by publishing hash numbers in The Times

this same idea of notarizing these events can be useful in contract disputes as well

Chris Wray:

This is an idea to borrow for civil proceedings too - we mentioned contracts that are unenforceable - you will never recover your costs. I you're above the small claims retail level but below a full arbitration these could be useful

this will trim cost and time, but you will still end up with witness statements etc, but for lower value disputes at earlier stages there will be verification

Vinay Gupta:

everything something goes wrong in trade, it adds risk, and reduces GDP - using no treaty mechanisms can grow global trade again


I was pleased to hear Rob bring up the subject of culture. We had an informal memorandum of understanding with an italian company, but the lawyer destroyed the trust through cultural difference

what are we going to do with all these other jurisdictions?

Vinay Gupta:

we dont knowwhat it means to have a global trade norms

Chris Wray:

humans don't like to think what happens when thinsg go wroing - contract law has been a class of people trained to do that - systems thinking is a better apporach

Kevin Marks:

why would you use a public blockchain for notarization of documents over a git repository?

Chris Wray:

the hashes themselves can give away information, so rarely this would be on a public blockchain.

Jeremy Goodwin:

I'm Jeremy Goodwin CEO of this session is less about the legalese and legal mechanics, and more about practical globalisation

How can blockchain tech support simpler creation of agreements and resolving commercial disputes

international trade depends on making contracts between different jurisdictions - GATT and WTO trade agreements have enabled this

seamlessly executed smart contracts will disrupt not only contracts but interational law too

the global economy is about $75T this year, growing at 3.1%

Blockchain Macroeconomic potential is high - the WEF predicts 10% of global GDP on blockchain by 2025

The WTO predicts a 15% tech trade boost by 2035 - 55% in China

the biggest contribution to the financial crisis was lack of transparency of exposure of companies to each other - a blockchain approach could have made this more trasparent

there are sovereign blockchain experiments going on in many countries, including multiple in Africa

Blockchain is resisted by contract law compliance, trade protectionism, customs brokers, and big companies opaque supply chains

manufacturing is a $12T global industry - it is growing in California again now

Manufacturing supply chain problems include exclusive relationships, centralized asset control, and RoI models for CapEx that punish R&D

supply chain with blockchain has decentralized protection of assets, and ledger of Things, enabling tracing of supply chains - this should replace cloud Enterprise Resource Management

Deloitte 2017 said 58% of manufacturers were implementing Blockchain

Our MFG token is meant to incentivize manufacturerers to be more responsive when quoting, and when redesigning

you ask them about blockchain and they understand blocks and chains; you ask about bitcoin and they say "isn't that a fraud?"


If you are launching a token sale, won't speculation make the token volatile so making it less useful for manufacturers?

Jeremy Goodwin:

we're only selling tokens to manufacturers and suppliers

Vinay Gupta:

there's a lot of capital tied up in inactive machines - any idea how much?

Jeremy Goodwin:

its not just that - they may have insufficixxxent operators too. In general it is running at 50% capacity, and they lack trained personnel

Michael Mainelli:

I'd like to talk about the economic impact of smart ledgers on world trade

Z/Yen is the city of London's leading commercial think tank - where finance meets technology

I've been putting computers into technology for 34 years so putting lipstick on the pig of a retail bank doesn't interest me

i built my first distributed ledger back in 1995 - I found some as far back as the 1970s

the block chain example of the distributed ledger is not a particularly good one, so I want to define things more carefully

mutual distributed ledger is a record of transactions shared in common and stored in multiple locations

mutual means shared in common, owned by a community - the internet is owned by nobody it is mutual in that sense

when people in business say that they haven't done mutual things before I point out email

we do a lot of research to back up my prejudices - just brought out a report on the impact of quantum computing on mutual ledgers

think very hard about the data you are writing out if you don't have a quantum resistant approach to it

what we have here is just time stamping - you can put a token on it if you like

it's a multiorganisational database with super audit trails, but it is a pretty crap database, so everyone using it will put a buffer database in front of it

normally the central third party keeps the ledger - now we can reduce the switching costs of the central party

smart ledgers: executable code is data too - executable code in a ledger is useful without tokens too

metrognomo is our time stamping engine for the channle islands

we do a lot of recording of clinical trials in a ledgers - 15m clinical assessments last year - a timestamping engine so that people don't cheat

I'm really excited about areas like identity - identity, documentation and information exchange

until last year I ran a commercial vessel - it involves a lot of moving of bits of paper, and it being questioned by people is what their job is

if you need 40 pieces of paper to move a boat, and the number of times they get bounced back and forth ends up costing a lot

then some banker comes in to show a distributed ledger to move money, and if it ever works, it goes up in price because of the transaction costs

in Geneva I saw Ripple say that a team of 8 engineers had taken 8 weeks to move $8 for $3 - and a chap from VISA asked 'you do know that we do that millions of times a day for 3ยข?'

Anyone who says they they are giving you database write for $35 they are an idiot

we did 25 billion transactions a day on our test rig - it could scale up to 1 trillion a day

We have a report on the Economic impact of Smart Ledgers on World Trade coming out in April

we're doing some work with the Singaporeans - we have about 500 companies on our exchange - they created an unowned ledger on our ChainZ software

It's called fasttrack trade - it's like Alibaba but decentralised - on a smart ledger

we have a distributed ledger of record at the bottom - you still need a database on top of it

there are 3 mutual levels - at the base there is a mutual distributed ledger protocol; next level up is business process, and that simplifies transactions; above that identity, documentation and agreement reduces uncertainty

we can see about a $25bn saving globally, but divided by the planet its not much, and compared to the cryptocurrency valuations, very low. I'll leave that with you.

we also found that there could be maybe a million new jobs, spread across the planet. So I've been unable to find the mega blockchain revolution that people keep selling

Richard Poulden:

I represent Black Swan capital, and our motto is "those who say it can't be done should get out of the way of those doing it"

Matthew Schulte:

now we do a lot of accounting in dull things like dollars and pounds, we will be able to do a new form of accounting that shifts what me measure

Michael Mainelli:

I think the impact in the developing world is going to be key - I remember when everyone was discussing how to cable Africa, when mobile phones came along, but what we didn't realsie that ti was for banking until M-PESA happened

Richard Poulden:

we had one asian country where farmers wanted to trade things forward - sell vegetables to restaurants ahead of time, whcih is perfect ledger application


When everyone has access to a phone that enables the creation for financial projects like M-PESA - every payment in Africa is routing through 2 new york banks - we can enable more people to benefit

Dan Bates:

I've been working in rural india and africa bringing renewable energy to them, and we ahve a blockchain driven smart meter that si addressable on a cellphone

M-PESA has about 70% of the transactions in Africa - the smart meter becomes a base layer that opens up the payment channel

Jason Brink:

one of the things that makes these powerful in the long run is that they are based on voluntary participation

also they can spread across national borders more easily

Michael Mainelli:

Matthew, you're trying to do an alternative thing - what do you think about payments?

Matthew Schulte:

there is a lot of activity that might be fostered when we can do micropayments - and also blockchains are not fully distributed - they need power and access to the global ledger

I agree fully with the point about mutual and mutuality - I like being able to play different games with different people - a user centred architecture, not application centred

Richard Poulden:

The payments going through New York is a dinosaur that will have its head chopped off - Bob Diamonds Atlas Mara is consolidating bad African banks to transform banking

if you have a pan-African bank finaincing African gtrade that will be a big winner

Jason Brink:

I live in Thailand most of the time, and these banking fees are crippling to small farmers, and micropayments matter there

Matthew Schulte:

Instead of payments, I'd like to use recognition - we can recognise these kinds of behaviour and record that - it moves from an industrial situation to a community one

Michael Mainelli:

I want to know what my net is, and tokens don't get that. For insurance to work we need pooling, and they don't add up

what changed your mind today?

Richard Poulden:

transaction speed and transaction costs matter - none of the cryptocurrencies are currencies

also people said 'well this won't work on a public blockchain' - so wjere are you goign to get a blockchain from


until the blockchain space has a real application that changes peoples life, it is still a big ponzi scheme

when we talk to communities in india and africa, they don't want to hear about currencies, but how blockchain can help

Matthew Schulte:

there was a conversation about trustful vs trustless - tust sin't liek that - we're moving where trust is held, but it is always contextual

if I'm eating in a restaurant then the trust that I want to know about is different if I am ordering lumber or booking a baly sitter

I want to see more trustful systems rather than trustless ones

Dan Bates:

I am seeing a lot of ICO things that are clearly nonsense - at the end of this there needs to be a financial underpinning to the eventually

we're going to run into this nuclear winter of blockchain will all these crappy companies and I want some survors with real businesses

Jason Brink:

we need to pay attention to the fundamental non-monetary uses of these blockchain technologies - identity and recordkeeping have deeper implications

look ats htm as tools for moving information

Michael Mainelli:

I'd like to compare this to my IV theory of technology - no new technology blasts anything out of existence - it's like ivy winding around a tree until the tree goes

it's what this enables us to do that we are not doing rather then emulating the things that already exist

in April 16-18th 52 leaders of the Commonwealth are coming to London, and a lot of their agenda is about Distributed Ledgers - if the trading systems of the commonwealth can pick this up it would be very interesting

Jeremy Silver:

about a quarter of you have been here before - let me explain a bit more about us.

Digital Catapult is part of a network of 10 Catapults that are meant to bridge between invention and ingenuity to implemntation

we run programs with companies large and small to drive adoption of these technologies

we focus on the creative industries and on manufacturing

Think how disrupted the music industry has beeen of the past few years, and see how that might apply to other forms of industry

we try and build spaces of collaboration for companies and startups to apply new technologies

one example is that in the games industry you often get small groups collaborating on things, and then some of them take off - so we use the blockchain to attribute their contributions

as part of the system we enabled dispute resolutions so when companies break up they can decide on distribution of the assets

we have 3 kinds of people - Developers, Business Analysts and Program Managers to keep us all in order

Vinay Gupta:

Tomorrow there is an event in Cambridge - Blockchain: Rewiring Governance at the Judge Business School

it's at

Rob Knight:

I'd like to thank Digital Catapult for making everything run very smoothly today