Featured Talks & Essays

Posted by | essays | No Comments

Cern Open Day Talk

Lift Conference Talk

Design Matters ( most recent essays are in this publication on Medium )|Money & Entropy (on Medium )|The UX of Banking (on Medium )| The Real Meaning and Future of Apple’s Mantra: Designed in California ( on Gizmodo )| How to Respond to Legal Threats with Cute Animals (on Gizmodo )| There is no such thing as invention ( on Medium ) | The Long Tail is Wrong ( on Medium ) | Lessons from a prelude to Pinterest (on Gigaom ) | The Path to Pinterest, visual bookmarks and grid sites (on Gigaom ) | Use Case Study House #1 – A house designed like a web application | Internet Pawn | Tech Crunch | The Big Apple | Demystifying-Demistifying | Tim Berners-Lee. Confirming The Exact Location Where the Web Was Invented | Ten things that defined the noughties | Social networks and creativity | Imagining the iPad | Susan Boyle and the second Great Depression (on Smashing Telly ) | Information and Evolution | What Comes After Lofts and the Suburbs | The Joe Ades Myth Deconstructed | Goodbye Dubai ( on Smashing Telly )

More talks:

Is the conventional bank model broken? Financial Times: Camp Alphaville, London
Beyond WWW, the cultural impact of the Web CERN, Switzerland
The Semantic Web SXSW Conference, Austin
Revolutionary Search Technologies SXSW Conference, Austin
The Post Spider World O’Reilly P2P Conference, San Francisco
Using XML for News Aggregation & Delivery WWW conference, Amsterdam
Microcontent SXSW Conference, Austin
Working Out of the Box Archinect
What mistakes to Avoid in Web 2.0 (I was so wrong – cringeworthy)

Venture Design and the Anthemis Foundry.

Posted by | anthemis | No Comments

Not just another incubator

Creating an incubator within many companies, today, is no more news than creating a marketing department is. In fact for our type of business, it is a given. So the launch of a financial services incubator at Anthemis, the Anthemis Foundry, isn’t news—the news is how we’re going to create companies and products.

Ecosystem fit and Venture Design

Most accelerators and incubators are about spotting or putting together an exceptional team with a great idea that addresses a large potential market. You need both of these, of course, but to solve the single biggest challenge of any startup, you need to solve the chicken & egg problem: How do you capture the potential market opportunity from what’s newly possible here and now?

This is what we call the ecosystem fit. Because Anthemis is all about ecosystems fitting together, it’s a natural approach for us. We’ll focus on three areas to build companies: the team, the eventual market opportunity, and the ecosystem fit – the starting point for the path to capture the opportunity. The ecosystem fit will be the variable that differentiates our focus, and design-based company creation, Venture Design, will be our differentiating process.

A previous ecosystem fit

Previously, I co-founded an incubator in San Francisco where I led the project that became Yelp. Having researched the space commonly referred to as ‘local’, which consisted of online classified services such as Craigslist, we realised that an online version of Yellow Pages was a larger potential market than Google’s then ad-revenue one, so Google would be forced to look at it. That was the market opportunity.

We then looked at the potential for an existing ecosystem fit, i.e., how existing organizations could possibly get from A to B to grab this opportunity. We realized that incumbents’ attempts at online versions of Yellow Pages would fail because they were not based on the new opportunities that were possible online (they did not even have search capability). Having written software to analyse Craigslist and then sifting through an entire month’s worth of local services classifieds, we realized that newer companies such as Craigslist also didn’t have the opportunity to get from A to B because their local services were riddled with covert ads for prostitution, so there was no way they could try to monetize it.

When we began to think how we might solve it, we looked at local review sites which were then nascent. It was clear that the area where Yellow Pages made most of its money (ads for doctors and dentists and plumbers) did not engage people the way that reviewing things like pizza parlours did. But if you built a big enough community of restaurant reviewers, you would eventually reach critical mass and get the ones for the dentists, too. The market opportunity was the reputation of all businesses, and the ecosystem fit was the pizza-joint reviews—the current opportunity that would get us there.

From market opportunity to people ecosystems

Human beings have a tendency to create god-like figures to hero worship. A celebrity culture makes things simple to understand and is a more romantic notion, i.e., the idea is that people like Mark Zuckerberg are unique. This idea is compelling, but provably false; Zuckerberg is extraordinary, but not unique. If you create an ecosystem with a particular opportunity, then a suitably talented person will find it.

This fallacy of unique people is demonstrable with two examples:

Example 1 – simultaneous invention: If true invention required unique ability, then the surprising number of examples of simultaneous invention would be statistically impossible. Evolution, the telephone, the equivalence of mass and energy, calculus and the vaccine for polio were all discovered separately by two people. The key word here is discovered. Inventions weren’t created out of nowhere but were new things spotted by suitably bright people building on the state of the knowledge ecosystem at that time and its history. The telephone couldn’t have been invented by anyone before electricity was discovered.

Example 2 – spooky connectedness: Multiple ‘unique’ people being in the same place at the same time is illustrated in 1927 when physicists attending the second Solvay Physics conference sat down for a group photograph.


The result was a definitive who’s who of 20th-century physics, including Erwin Schrodinger, Werner Heisenberg, Paul Dirac, Max Planck, Marie Curie, Albert Einstein and Neils Bohr. The revolution that replaced Newtonian physics was created by people who knew each other personally, thus forming part of an ecosystem. This could have been a coincidence, but it happened in other sectors, too:


In August 1958, Art Kane, a freelance photographer for Esquire, asked if some jazz musicians would turn up for an impromptu photograph in Harlem. Fifty-seven of the world’s most famous musicians came, from Art Blakey to Count Basie, Dizzy Gillespie, Charlie Mingus and Thelonius Monk, creating one of the most famous group photographs in history. They just happened to all be around at the same time. The reason is they all knew each other, jazz was an ecosystem.

Today there is an ecosystem of internet entrepreneurs, and most of them are in the San Francisco and wider Bay Area. Their connections are a complex mix of social and professional, with rarely more than one degree of separation between them. Caterina Fake, who founded Flickr with her former husband Stuart Butterfield, who founded Slack, used to date Evan Williams, who founded Twitter. These are seemingly accidental personal connections that predated their entrepreneurial success.

The above relationships involve many people, not just two, by multiplying the chance of a chance of a chance. They are even more statistically impossible than simultaneous invention until you view the world not in terms of unique people and special ecosystems, but in terms of special people and unique ecosystems.

Strategic importance of digital ecosystems

The digital ecosystem is important; very important. We are undergoing a change to society that is now proving to be as fundamental as the Industrial Revolution. It’s effects are permeating modern life and extending into politics as the social pact between those who own the means of production and the ‘workers’ is threatened by ‘algorithmic’ platforms where the means of production and the workforce (computers) are either one and the same (internet companies employ 300 times fewer people per dollar market cap than industrial ones) or where workers are no longer employed in the traditional sense (40% of Americans will be freelance by 2020).

Europe led the Industrial Revolution, and although it culminated in the US with the assembly line, Europe created its share of large industrial companies. Along with the American companies of Ford and General Motors, there were Italy’s Fiat, France’s Citroen, and Germany’s BMW and Mercedes. With the internet, however, this has not happened because internet platforms, like the internet itself, straddle borders. And so, in both America and Europe you have Google and Amazon as market leaders.

But the situation is particularly dire for Europe. Very large numbers start to sound the same in newspapers, but $1 billion is a lot less than $100 billion and few people have noticed how far behind Europe is. The combined value of the top three internet companies by continent is The Americas @ $0.75 trillion, Asia @ $0.5 trillion, Africa @ $50 billion, and Europe @ $25 billion. Europe is behind Africa technically, and although it has billion-dollar internet companies, it doesn’t have $100-billion ones, which would have emerged if the tech ecosystem had been globally competitive.

Potential for a fintech ecosystem

Anthemis has been at the centre of technology-led financial services revolution long before the term “fintech” existed, and its model, even its name, is predicated on the idea of connecting and stimulating an ecosystem of modern, digitally native, financial services. The opportunity exists for a sector-specific financial services internet ecosystem outside of the US to dominate, and Anthemis, with our base in London and global reach—is perfectly positioned for this ecosystem. This opportunity exists for three reasons: (1) money is a universal language that does not need a marketplace of 300+ million American English speakers to dominate; (2) London, more than New York, is at the hub of a truly global approach to finance; and (3) regulatory arbitrage opportunities, which have been seized with Silicon Valley style libertarian zeal when launching companies such as Uber, are trickier with finance.

Process to capitalise on ecosystem opportunities: Venture Design

So now is the time to capitalise on the potential opportunities that exist to deliberately create financial services companies from scratch, and now is the time to use our formal approach: Venture Design. This approach is the deliberate realization of opportunities by competent people rather than the passive search for unique entrepreneurs who have found them; it is manufacturing companies rather than spotting them. Venture Design is only possible because people are not unique, but ecosystems are, and because a design approach means you can create novel things on demand.

“Venture Design is only possible because people are not unique, but ecosystems are, and because a design approach means you can create novel things on demand”

Few people would say that architects can’t create works of art because they are creating building designs on demand, but the process of company incubation can seem cynical and factory-like because it is premeditated. However, creating companies in an incubator isn’t necessarily of lesser creative value or quality than companies created independently by entrepreneurs as long as there is a creative process.

Anyone can be creative some of the time but a creative professional, such as a designer, can do it on demand. The trick is that designers are not actually creating things from scratch; they are seeing newly emerging patterns of opportunity and novelty and capitalising on them, that is, they have deep knowledge of a particular topic (such as Anthemis has with digital financial services) to enable unique insights. When this creative process is extended to the act of creating new companies, rather than the sorts of things we normally associate with design such as products, graphics and buildings, the formal process of Venture Design can evolve and it can be repeated. In other words, you can design high-quality, innovative companies on demand.

Venture Design is what we will be doing at the Anthemis Foundry. Because Financial Services is an ecosystem where a market opportunity for digital disruption has been visible for some time, but where path to get there has only recently emerged, now is the time to do it.

Eris, a Software Stack Including the World’s First Blockchain Application Server.

Posted by | Uncategorized | No Comments

As part of a new project which we will be launching at Anthemis I have been helping a groundbreaking startup called Eris which launches today.

Eris is a software suite including the world’s first application server based on the technology behind bitcoin – the blockchain. It is designed to enable any kind of transaction between parties by enabling trust in a neutral framework.


The importance of creating trust via neutrality is something Eris’ CEO Casey Kuhlman has specific experience of. As this piece in the New Yorker relates, as a U.S. Marine in Iraq he was the person that had the presence of mind to replace the US flag with an Iraqi one, when the famous statue of Saddam Hussein was pulled down. After leaving the Marines Casey became a lawyer involved in human rights and worked in Somaliland to create a legal framework. This is the maverick background which lead to Eris industries, a story that sounds almost too far fetched to be true but creates the unusual context that is often the basis of the most innovative companies.

What does Eris consist of?

Eris is a full stack of software to enable you to build applications using blockchain technologies not just currency transactions ones but any kind of self-executing contract, financial or legal. These are known as Smart Contracts.

There are 4 components in the Eris stack: 1. the deCerver, the world’s first blockchain based application server that can use any chain design including 2. Eris’ custom, smart contract enabled (i.e. any transaction, not just money) blockchain design, Thelonius, that can be used to create chains for groups of people or corporations or specific applications, 3. a package manager to manage multiple contracts within a single application and 4. a contract markup language that can be used to bind real world legal contracts to smart contracts (which, as software scripts are necessarily logical and unambiguous unlike most real-world legal ones).

So what is it?

Its a fancy new type of database.
The blockchain is a fancy database, except that there is no need for a master and slave, just peers. Unlike the first generation of peer to peer systems 15 years ago which resulted in services such as Skype and Napster, blockchain based systems are truly decentralised (no master/slave at the database level) rather than merely being distributed. The Eris stack is a kind of ‘mesh’ database.

With an self-governing audit trail.
But this goes one step further than being a mesh, because the verification of each block means that a blockchain database maintains state and a transaction history. A normal database has a state in time and an ancillary log file, in a blockchain one, the data is within the log file itself and the log file is completely secure from tampering by achieving consensus via cryptographic means. It is a mesh database with a self-governing audit trail.

Capable of triggering scripts based upon rules between parties – contracts.
But it gets better. Just like databases can have triggers, pieces of code that can execute within the database, and just as an object-oriented program can instantiate code with data together, this is a blockchain where the blocks contain not just monetary transactions but any kind of transaction. This is achieved by allowing scripting in the blocks of the secure blockchain, in the manner that blockchain designs like Etherium allow.

What kinds of use cases could Eris applications serve?

In some ways the main focus of Eris is to create beautiful tools that enable developers to create applications for new and interesting use cases, however in developing the system, 3 stand out:

1. Fintech
2. Legal tech
3. Internet of Things

1. Financial contracts are about quantitative elements (money) and so lend themselves naturally to smart contracts which have to be unambiguous. The promise of the blockchain approach is to play a key part in the ongoing replacement of the processes and plumbing of the financial services industry with Internet based ones. This has far reaching consequences and is why fintech has become a very active sector recently (and the area I now devote all my time to).

2. Legal innovation based on Internet technology is not as advanced as financial services and the number of people who are both lawyers and coders is vanishingly small. But the opportunities are potentially just as exciting and the Eris team consists of some of the world’s leading experts in this area. A blockchain approach allows for consensus to be achieved without a leader or intermediary and for that consensus to be protected from hijack by parties interested in corrupting it. This allows for all sorts of things to work better from voting in an election to complex contractual relationships between multiple parties, such as on a construction project. Unlike financial contracts which are inevitably dependent on numbers,legal contracts consist largely of words, so their obligations often have some degree of ambiguity which makes them difficult to bake into an auto-executing smart contract script. Eris’ approach is particularly pragmatic here as it allows for a mix between smart contracts and real-world legal ones, without which there would be no opportunity to create useful applications.

3. Perhaps the most exciting opportunity for Eris is the least obvious. The next phase of the Internet, the Internet of Things will be where ordinary physical objects are connected to the Internet and often communicate with each other independently of humans (from components in a smart factory, to sensors in homes). The result of this will be a complex network of devices communicating with each other automatically, via software APIs and the role of smart contracts to mediate this communication could be pivotal.

Minimum & Maximum Viable Civilizations,

Posted by | Uncategorized | One Comment

The physical constraints on the design of civilizations and how this relates to the Fermi Paradox.

One of the features of any civilization is information transfer. In fact, this is the principal feature that can be abstracted to create a mathematical foundation for the basis of most of the activities we associate with civilization: groups of people communicating with each other; trading; accumulation of knowledge and the end results of these things, namely culture and technology.

How might the physical limits on the ability to transfer information, (speed of light and quantum effects) affect the maximum and minimum possible sizes for civilizations and therefore the search for Extra Terrestrial Intelligence — SETI?

Current SETI is often based on two ideas:

1. that really advanced civilizations will grow bigger and capture larger and larger energy sources, from the home planet to the entire output of a star, via a Dyson sphere, say, to eventually colonize an entire galaxy. These are referred to as Kardashev Type I,II and III civilizations respectively. This idea implies we may be able to detect the big ones.

2. that the shift from a Type II to a Type III civilization can happen quite rapidly, because self replicating machines could spread though a galaxy. So if big ones are likely, there should be some, given the age of the universe and the fact the we are here.

Neither of these take into account the fact that individual members of civilizations need to be able to communicate with each other or they aren’t a single or evolving civilization, by definition.

Freeman Dyson has estimated it would take us 200 years to reach type I status at current growth levels, and with a modest growth rate of 1% per annum, Nikolai Kardashev estimated that it would take 3,200 years to reach a Type II status, and 5,800 years to reach Type III.

Given this growth rate and the following other parameters:

(a) the (in my opinion quite reasonable) belief that living things are a naturally occurring and inevitable form of low entropy order that efficiently channels the energy of an open system, towards waste, like the structure of eddies in turbulent flow.

(b) the estimated number of planets. Current estimates are around 100 Earth like planets alone, for every grain of sand on every beach and in every desert in the world.

(c) the age of the universe. 14 billion years.

…the original formation of the Fermi Paradox, ‘Where Are They?’, becomes ‘WTF are they?’.

Challenging the first assumption: that an advanced civilization will get bigger.

We are a long way off being a Kardachev Type I civilization in energy terms. We use around a millionth of the tiny proportion of the Sun’s energy that ends up on Earth, partly in the form of stored solar energy from fossil fuels. Yet we are already in danger of making our planet uninhabitable for us.

Our current issue is primarily CO2 emissions, but it could be that increased use of what are currently perceived to be sustainable energy sources turn out not to be, because we will have an ever increasing dominance on the rest of the living eco-system, if we continue to grow.

So what if we shrank, instead?

Our attempts to create a sustainable eco-system are largely based on efficiency and the logical extension of how to grow, say our population, with greater efficiency is for us to get smaller. If, at some point we replace ourselves with evolving, self-replicating machines, why not make them expand by making themselves smaller at each iteration, to colonize by contraction within the earth rather than expansion through the galaxy and through greater intelligence density rather than extent – more bits per atom.

Atoms are small, using the staple grain-of-sand measure, a single grain of sand contains more atoms than there are stars in ten million galaxies, so thinking machines could certainly get much smaller. But you need a lot of bits to store and process the information that we and our civilization requires, so going small may not be as good a solution as Kardashev style expansion is, long term, unless we can go a lot smaller than atoms.

As it happens, this potential growth via contraction to the smaller scale is gigantic, bits are theoretically limited to the planck scale and this is truly tiny. We sit almost exactly half way between the Planck scale and the size of the universe. There is theoretically as much room to colonize via contraction as there is via expansion.

But there is a limit to how viable a small machine could be in terms of physics. Even at the atomic level, quantum effects, and in particular, Heisenberg’s Uncertainty Principal, become significant. This says that you can’t know the position and momentum of something absolutely, because at the minimum you have to use light to measure it and at small scales, photons affect things as large as atoms. The Uncertainty Principal limits the scale at which effective measurement is possible and effective measurement is the key to information exchange, which at the beginning of this article, I suggested was the key abstraction for a mathematical analog of civilization. Civilization requires information exchange, and if you can’t exchange information reliably between one place and another, they cannot remain part of the same civilization.

The Uncertainty Principal creates an absolute Minimum Viable size for a civilization and it is larger than the atomic scale, i.e. ‘Reverse Karadashev’ expansion is limited to a size much, much bigger than the smallest things in the universe. It is analogous to Kardashev types not being possible beyond the size of the solar system.

Not many people would argue with this, that vastly smaller than sub-atomic machines are physically impossible, let alone sub-atomic civilizations. But nobody has challenged the physical constrains for the Maximum Viable Civilization and how that might affect both the number of potential civilizations and the ability to detect them — SETI.

The physical constraint at the large scale is also key to information exchange. At large scales, instead of message fidelity, the problem is speed of transfer. Information transfer is limited by the speed of light, and at the galactic scale it is incredibly slow. Our galaxy is an average sized one and it is 100,000 light years across. The average Kardashev type III civilization would take up to 200,000 years to get an answer to the question ‘are you there’ from one of its own members. Not knowing whether your other members exist or not, means that the speed of light suggests that a Kardashev type III civilization is impossible, on average, for members of a civilization who live for less than 200,000 years.

Several people have suggested that interstellar colonization would require ultra slow ‘metabolisms’ so that communication over large distances would appear realtime (remotely driving a rover on Mars is impossible in realtime, for humans) and organisms or any object which decays through use would be able to extend their lifetime. But this would mean that they would have to create ultra-stable artificial environments, where rivers didn’t change course or even meteorites strike to avoid having to react to things which changed much faster than their metabolism (or more accurately their brain cycle, or whatever) dictated. There is possibly a limit to this stability, and this implies a Minimum Metabolic Rate: MMR

Civilizations continue to build knowledge and culture over time, through the exchange of millions of bits of information, as humans do, every day. A Maximum Viable Size for a civilization may mean that a minimum number of bits of information exchange between members required to constitute a civilization which evolves as a whole is a relatively large number of messages, certainly much more than an endless series of pings of ‘Are you there? Yes. Oops, I’m dead.’ Call this the Minimum Messages per Lifetime: MML

Putting these parameters together (speed of light, age of the universe, MMR and MML) allows for a Drake-like equation where numbers can be plugged in to estimate the Maximum Viable size for a Civilization.

Say MML is 70,000 messages, this means that 70,000 messages would need to be sent back and forth (140,000 trips) across 100,000 light years for an average Kardashev Type III civilization to be possible. i.e. the lifespan of a single individual would have to be 14 billion years, which is larger than the age of the universe.

Given the age of the universe and the average size of a galaxy, for Kardashev Type III civilizations to be possible, MML must be less than 70,000.

For the messages to appear realtime, the rate of messages would need to be roughly correlated to the metabolic rate. Say the minimum MMR is one message per 714 years, this would mean that if MML was 70,000 the maximum sized civilization would be 100,000,000 light years across. i.e. a Kardashev III civilization would just be possible for the average galaxy, any faster then it wouldn’t.

Given the maximum value for MML, a Kardachev type III civilization is only possible if MMR is less than around 1 per 700 years.

If we start to estimate the viability of these parameters we can estimate the Maximum Viable size of a civilization and the possibility of Kardachev Type III civilizations.

Apple iWatch Design Predictions

Posted by | Uncategorized | One Comment

The First Internet Company In Shoreditch

Posted by | Uncategorized | 2 Comments

In 1994 I started the first Internet company, Realtime Anywhere, in Shoreditch, what is now, 20 years later the center of London’s ‘Tech City’. What was London’s Internet center like then and how has it changed?


The reason I know Realtime Anywhere was the first is that there were almost no Internet companies in Britain and almost no technology businesses of any sort in Shoreditch, where all of the people involved in creative industries there knew each other. After the property crash of the 80s, when Interest rates hit 18%, light industrial units around Hoxton and Shoreditch, were often derelict, rents were cheap and the area boasted the largest proportion of artists of anywhere in Europe.

I was an architect at the time and several architect friends had moved there into quirky, unusual places. A friend, Dan Philips’ place was so big you could ride a bicycle in it and one local resident lived in a signal box on a disused railway line that you had to get to by ladder. I found an empty and very cheap, 3 storey warehouse in Cotton’s Gardens, near Hoxton Square for roughly the price of a parking space today (40,000 pounds) and changed the usage to live-work. This was a new planning option which had been brought in as services like desktop publishing had replaced the partially toxic light industries such as printing that circled London’s financial center. Redevelopment of Liverpool Street station had pushed the boundaries of city workers’ lunchtime excursions northwards, bringing a cluster of seedy strip clubs, or more accurately ‘strip pubs’, most of which have disappeared apart from the infamous Brown’s where bankers and East End gangster’s would mingle to gawk at lunchtime strippers who included the woman who ran it’s daughter.

There were no Shoreditch pubs or bars open at the weekend except the Bricklayer’s Arms on Charlotte St. and the ironically parochial Charlie Wright’s International Bar. Both were served by Lil, an orange-rinced barmaid from a Carry-On film. Despite the lack of fancy bars, shops or drinking club franchsies like Soho House, the atmosphere was surprisingly similar to today. Shoreditch still looks superficially grim and somewhat Dickensian despite the gentrification and it isn’t so much more colorful on deeper inspection, as creative professionals have displaced the more avant-garde who decamped to places like Berlin. The London Apprentice gay club was there in the 90s and the Blue Note was the first piece of nightlife to sit on Hoxton Square. In the mid 90s, just under the railway bridge over the Kingsland road, the first Vietnamese restaurant opened, now part of what could justifiably be called Vietnam-town.


Myself and Markham Darbyshire started Realtime Anywhere in 94 from my living room at 5 Cottons Gardens. His dad came to give us advice and when we said that we thought there was a business in creating websites for people it wasn’t obvious there was a market for that. We also thought domain names might actually have a value but we’d need money to host them (domains were free to register, but you had to have an IP address to host them and Demon Internet in Finchley were the first place in the UK that you could do this for reasonable cost). We registered anywhere.com as our domain and our company email address was anyone@anywhere.com.

We had dial up Internet and later ISDN. My friend Simon Perry started an Internet company, POW communications, down the street and he was the first to have a leased line with a then impressive 64kb connection. Simon kept an Internet diary at the time, noting down the first time he heard people mention the web in the Bricklayers’ (Bloody hell – can you imagine people are actually talking about the Internet in Shoreditch!), the first time he saw a URL on a billboard or read out on TV.

View Larger Map

We started working on more involved projects, a school friend Angus Bankes who had co-founded a healthcare technology company joined and we developed an Intranet connecting hospital staffing around the UK, a genealogy service for Scotland’s government records (with my father, who had even earlier Internet cred, having written possibly the first Internet book – The Electronic Mail handbook, in 1979) and some experiment projects such as a 3-d search engine. We then teamed up with Nick Denton, who now runs Gawker Media, and I later moved to San Francisco to create the first news aggregator, Moreover.

At the time, however, Shoreditch was considered a backwater, and when in 97 we were given the opportunity of an office in Soho, above Mr Bongo records in Poland Street, in exchange for doing their website, we jumped on the chance to move somewhere that would improve our credibility as a technology company.

Skunk Puppying: dealing with legal threats with cute animals.

Posted by | Uncategorized | 15 Comments

I registered designskunkworks.com, thinking it would be a good name for an Internet product company, not realizing that Skunkworks was a real, trademarked facility run by Lockheed Martin or that you can now threaten to sue people just for registering a domain name.

Having long been a fan of David Thorne’s hilarious email exchanges on 27b/6 and wondered if the technique could be put to use in a real word scenario. Big companies don’t do surreal. So when I received the nasty letter below from Lockheed Martin’s lawyers, I sent them a picture of a kitten.


Below is the email exchange that ensued:

Dear *****,

I’m happy to transfer this domain, since for ethical reasons I do not wish to be associated with Lockheed Martin, the manufacturer of cluster bombs and weapons of mass destruction, or any of the people such as yourselves who represent them in such a harassing manner.

The domain was bought in good faith that it was not a trademark violation, since skunkworks is a commonly used term. I do not host this domain and any monies being made from it are by the domain registrar, whom you can contact separately. Since I have no idea how to assign a domain please send detailed instructions.

Here is a picture of an 8 week old kitten taken by Sasan Geranmehr to take your mind off things http://en.wikipedia.org/wiki/File:Persian_Cat_(kitten).jpg




Dear Mr. Galbraith:

Thank you for your cooperation. In order to transfer the domain name to Lockheed Martin, we will need you to do the following:

(1) Contact your registrar, Domain.com, and request that the domain name be unlocked;

(2) Also request that Domain.com provide you with an authorization code to transfer the domain name (this may not be available for 60 days following the registration of the domain name);

(3) Once you obtain the authorization code, please forward it to my e-mail address as well as to **************@******.com;

(4) As the administrative contact for the domain name, you will then receive an e-mail from Corporation Service Company (i.e., ****************@*******.com) asking that you approve the transfer of the domain name to Lockheed Martin Corporation. Please reply to that e-mail stating “I agree” to the domain name transfer.

We look forward to receiving the transfer authorization code.




Hi *****,

I was a bit disappointed that you didn’t thank me for the nice picture of the kitten I sent you.

I thought it was a kind gesture since, after having registered a domain that I quite reasonably believed was merely a concatenation of commonly used English words, I was wrongly, distressingly and somewhat unbelievably accused of attempting to harm the business and reputation of a colossal arms manufacturer. I was also incorrectly accused of attempting to make money from this, when simple due diligence would have shown this to be false.

I’m further distressed since having done a Google search for Lockheed Martin to see if you were in fact acting for them or this was just some kind of weird practical joke, there were some really gruesome images of people with no eyes and disfiguring wounds under the title ‘Lockheed Martin Cluster Bomb victim’. I need to see some pleasant and reassuring pictures of cute animals to remove these horrible images which have permeated my subconscious.

So here’s the deal. I’m out of pocket $9.99 for the domain you want me to give to you and doing this transfer to gift it to you is a considerable hassle vs letting the domain lapse, as is, without me ever using it (I have cancelled automatic annual renewals and will never host any website there), which I agree to do unconditionally. I also doubt I would win a fight against someone with nuclear weapons with my Swiss Army Knife.

If you send me a link to a picture of a really cute puppy, I’ll transfer the domain to you.

Best Regards



Dear Mr. Galbraith:

We appreciate your cooperation. Our client is willing to reimburse your costs for registration of the domain name if you can provide us with proof of same (e.g., receipt, invoice, etc.). We will also need you to complete the attached W9 form to process any reimbursement costs. Thank you.




Hi *****,

Sorry for the delay, I was in London, which would have been nice had it not rained cats and dogs.

Speaking of which, since the costs of me providing you with what you need to process the $9.99 reimbursement you offer would exceed the amount offered, and since, without doubt, the costs for your client to process this would also exceed the amount. Further, since the costs for yourself to organize this reimbursement would also certainly exceed this, it would be both a waste of money for all parties concerned and merely a symbolic gesture.

I am sure we agree on not wasting money and I am happy that you are interested in a symbolic gesture. The picture of a puppy will be much cheaper and fill me with the joy that money simply cannot buy. It will more that recompense for the tedium of following the steps your client would like, beyond merely not doing anything with the domain DesignSkunkworks.com

I look forward to receiving a wonderful picture of a puppy (as long as it isn’t a Chihuahua, or any other dog that is smaller than a domestic cat when fully grown) after which I will gladly go through the detailed instructions your client would ideally like.

Best Regards



Hi *****,

I didn’t receive a response from my last email of July 24th, with the
suggestion of a settlement using puppy pictures rather than cash. I
thought I might, since your first letter contained all sorts of
terrifying crypto-intimidatory legalese like I imagine you get when
you haven’t returned a library book for 36 years in a Kafka novel,
demanding “steps” and “written assurances” etc. with a deadline of
“August 6th, 2012”.

It’s now September and I’ve received less than one puppy image.

I understand that puppy picture payment must be a more demanding
logistical problem, due to its unorthodox nature, for a corporation
like Lockheed Martin in comparison with Byzantine legal procedure or,
indeed, the manufacture of thermonuclear devices. I can comprehend
that, it’s no doubt a serious business and gravitas is important

Gravitas, isn’t important to me, however. I genuinely believe the
world would be a better place if eminently solvable legal disputes
were resolved by the ludicrous ritual exchange of somewhat cloyingly
saccharine pictures that nevertheless made people whose personality
has been subsumed by institutional groupthink melt like grannies
huddled round a new born.

I’ve invented a new term for this, Skunk Puppying.

Accordingly, if I haven’t heard from you by September 17th 2012, I’ll
assume that you are taking this matter no further and I’ll have to get
a copyright free or Creative Commons (CC BY-SA) puppy pic, elsewhere.

All the best



Skunkpuppy competition.

I thought it would be good to to a Skunkpuppy logo for other people to use when skunkpuppying. If you want to reward people who send you nasty letters with pictures of cute animals, you would be free to use this under a Creative Commons license.

I was thinking something along the lines of a cartoon dog wearing a superhero outfit with his fist in the air.

My friend Keith had a stab at some ideas already:


Submit your entries in the comments.

5 principles of invisible web design

Posted by | Uncategorized | 3 Comments

The term design is associated with visual things, we design magazines, products and buildings but not music or food and this visual association has carried across when people speak about web design. But Like food, it’s as much about getting the right recipe or the right use case, flow or need – Invisible design.

Here would be my picks for 5 design principles based on invisible web design:

1. Design for beneficial feedback loops
Beneficial feedback is the opposite of virality. A virus spreads but is harmful, something which is beneficial spreads precisely because it is valuable. Don’t design things to get more users, design things where the more use the more value and therefore the more users. Focus on the product and the business will come.

2. Design for incremental value
Create something where the more people use it the more value there is for each of them. This creates community.

3. Design for the primary use case.
Do one thing and do it well. There is no scarcity of resources in Internet land so a secondary use case can be a secondary product. The scarcity is attention, design for the principal thing that will engage people.

4. Design the personal.
Small companies used to pretend to be big companies by appearing less personal .e.g. a switchboard that said “for customer service press 3”. Do the opposite. Design so that 30,000 people appear like 3.

5. Design recipes not visuals.
Achieve 1-4 with recipes not visual designs. A web design will have a specific flow and ingredients. How this looks will change based on the technology it sits on. But the recipe will be universal.


Use Case Study House #1 – A house designed like a web application

Posted by | essay | 19 Comments


User Experience House Design

Between 1945 and 1966 Arts and Architecture magazine commissioned a series of case study houses which have become icons of modernism.

Since I’m both an architect and a web product designer, I thought it would be fun to design a building the way that web applications are. Use Case Study House is a pun on the Case Study program and a reference to use case design which is fundamentally different from the way buildings are planned.

Web Design vs Building & Product Design

Web design has come along way in the last few years, since the rise of User Experience (UX) specialists, who are closer to architects in what they do than traditional web designers were. That is not to say that some web designers weren’t good UX people, it’s just that the perception of the role of web designer meant that they were marginalized into graphic design.

Again, this is not to say that graphic design is marginal, just that the web is not a graphic design medium but a product design one. This is why many corporations ended up with good looking, Flash-based websites that were largely useless in terms of doing the things a website is supposed to do. At the other end of the spectrum were the pure usability people like Jakob Nielsen who knew how the web worked but produced things that looked awful. A user’s experience is not just determined by functionality but by all sorts of things such as history, culture and perception, or nobody would use a qwerty keyboard or drive a speed limit breaking Porsche.

Because the web is relatively new, web design still has a way to go before it mirrors what architects and product designers do. For example, VP product is not a standard role in startups and where it is its usually less senior than the CTO, or VP of marketing. Much enterprise software and earlier dotcoms were not design driven cultures, so marketing or product marketing people would gather features and have the final say on what was passed to engineering. In architecture this system is only used for low design quality design & build projects.

For web design to move beyond User Experience to be more like Architect driven design, UX people need to be both more technical and more aesthetics driven. [ Update, February 1st 2011: What I should have added here is that UX should also be less about, brainstorming, interviews, focus groups or research – in architecture this constitutes preparation of the brief is often done by the client, and has little to do with the actual process of designing things. ] The role needs to expand from UX so that a ‘product’ person will actually design the architecture of the system (schemas, hardware and scaling strategy, and specify exactly what the product does), but from a designer rather than engineering perspective. I’ve been banging on a about this for years, but few people seem to relate to the problem, so this is not conventional wisdom and possibly wrong.

Somebody said (can’t remember who) that if the great 19th Century engineer, Brunel, was alive today he would have been a software designer. Software design is more interesting because there are things that haven’t been done being innovated all the time, its like being an creative engineer in the Victorian era, and so if people aren’t ready to design software the way architects design buildings, what about trying to design a building based upon web design methods – nothing complicated just a single diagram.

The diagram

The picture above is the end result. The title is a play on the Case Study Houses of the 1940s. Its not a UX design but a UX inspired one. Neither is it an engineering one – its doesn’t use UML or all of the very structured diagram methods 1. because they don’t map to this process and 2. because I don’t like them for these kinds of tasks. Things like UML work for very well defined rational scenarios but user focused design like a house or consumer software is rarely rational, its based on intuitive feel, so you can invent your diagrammatic language as you go along, mix metaphors and use things inconsistently, in order to get the message across to another irrational human being.

How its different

Web design is very linear, its all about flow and eliminating the niche, to get the bulk of people through a primary use case. Many architects tend to think of buildings as objects, the greatest ones, such as Frank Lloyd Wright, often thought about them as interconnected spaces but they focused on the spaces rather than the flow through them – this is analogous to looking at the stage set rather than the choreography. I’ve often wondered what it would be like to design a building like you would choreograph a dance – so that the end design was a picture of a person moving rather than the environment and where if that was sophisticated enough the environment would be defined by the person’s movement.

User Experience based design is more like choreography.


A specific example

I stayed in a hotel recently where the bedroom was connected to the bathroom by a small dressing room corridor. This arrangement really worked – clothes were neither in a ball on the floor on the bedroom or bathroom and the whole flow of getting up and going to bed worked like a perfectly executed dance or a good website, with everything where you wanted it and in the right order. So much so that I measured it (always carry a tape measure).

Why is this at all interesting?

Good design is often about things which seem mundane and obvious after the fact. An example of this on the web would be the ‘permalink’ – few people can remember that before humble online diaries informed people like the New York Times how to publish on the web, you often couldn’t share a link to a news story (it would be at a url like ‘topstory’ which changed when the top story changed). In the architecture world, most people don’t have dressing rooms or walk in closets, but this is less a function of cost and space than of design, most houses have corridors and cupboards. so in the flow based design above the bedroom module uses the same flow as was in the hotel. Added to this flow is a connection to laundry. In the UK in small houses many people put a washing machine in the kitchen, in the US laundry is more often in a basement and in Hungary the washing machine is usually in the bathroom. The last arrangement makes most sense, but it would be better to combine the US and Hungarian model with a flow which has laundry going from a space between washing and sleeping to a laundry area – e.g. with a laundry chute.

The design

The overall concept of the design is to separate the spaces into active and passive, and look at the day to day activity flow to determine how to subdivide and arrange them. Most rooms in houses are based on obsolete uses, hence the ubiquity of loft spaces – which were obviously based on re-using industrial spaces that were not originally designed to be lived in at all. The drawing room has merged with the living room which in turn has morphed into a TV room as the kitchen and dining room have also merged, giving more living space. This is what starts to move towards spaces which are defined less by activity than by type of activity and the type of environment. This, for example, is how traditional Arab courtyard houses worked based on prototypes in Baghdad – rooms were multipurpose and labeled according to the type of climate in them. A house subdivided primarily into active and passive spaces could have a very different feel for each: cozy, soft, dark for the passive, larger, bright, hard surfaces for the active one and this would in turn create different flows.

For a view of what this would look like in reality, I applied a watered down version of the active/passive subdivision of space to the barn we converted into our home, below are two pictures of it, to illustrate the point.

The passive space is, cube shaped wooden and windowless, while the active one is rectangular with hard surface and bright overhead rooflights and proportions scaled down from a banqueting hall which make it appear much larger than it really is.

Active space:


Passive space:


A new room type, the ‘Maker’ Space.

Going beyond this rough active/passive concept for the flow diagram for Use Case study house #1, I though it would be interesting to add the activities that often happen in the garage which has gone beyond the role of storing cars to being a place of creative activity throughout the US, everything from bands to Apple computer started life in the garage. I brought this space into the house and along side the main active space (the dining area) so that there was an activity space that was all to do with play and making things. A maker space.

Active zone:


The original bedroom flow has been created as a module which can be repeated multiple times and linked to the service spaces (which are all humid – kitchen, bathroom).

Sleeping module:


This might all seem pretentious – the end result is just a simple diagram and its by no means a complete design. But its a diagram that has a lot of work put into it to make it the right diagram, and that’s what makes a good design for a web application.

Internet Pawn

Posted by | essay | No Comments


For me, there is no startup that gloms onto the depressing zeitgeist more wretchedly than Borro which follows mockney Wonga into the foggy world of dodgy debt – the very thing that got us into this mess in the first place.

According to their testimonials, “borro’s service sits somewhere between corporate investment and pawn broker”. Exactly one Planck Length away from the latter, because in my experience, corporate investment doesn’t tend to involve leaving your wristwatch as a deposit. With its lending service, Borro wraps the world of technology, banking and the Great Recession into a nasty little ball, with a story so perfectly intertwined that it reads like contemporary Dickens.


Borro.com, Boo.com with two ‘r’s.

Pawnbroking is when money is loaned against something real but which can be sold quicker than a house if massive interest is not paid. Its in the 7th circle of debt hell, three rings below sub prime, where desperate people sell their granny’s ring to buy crack. It’s not a nice business, but in times like this, its a thriving one.

Things that you can sell quickly have to be reasonably desirable, so pawnbrokers’ windows are full of the best things that poor people have ever owned, giving a bizarre impression of status amongst the deprived. In a pawnbroker you can find gold jewelry and expensive watches and brands that indicate respectability. If you were to extrapolate this further, a ridiculously distorted caricature of a pawnbroker in a dystopian Gatsby would store fine wines, famous paintings and fancy cars. But in this cartoon, behind the mask of respectability would be something vile and predatory.


Montague Tigg outside the London pawnshop.

This Gatsbyesque image of diamonds & Porsche’s is exactly what Borro claim to have attached to pawnbroking and without the vileness behind. Sections on their site include diamonds, gold, fine-art and antiques, vintage cars and, unbelievably, luxury yachts. Items are stored in time-locked steel cages with ‘tremblers’ and 24/7 CCTV, except the yachts of course! In addition there is a section on their site about their values which outline ways they differ from the normal practice of pawnbroking.

There are a number of possible reasons for selling pawnbroking as upmarket, whether they work or not. You could try to make pawnbroking completely up-market with a more lucrative clientele or you could make it less downmarket using a few rich users as anchor clients to attract the masses or you could create the illusion that rich people are using the service to reduce the stigma which a user base of ordinary pawnbroker clients would feel.

So is it real?

According to the Wall Street Journal, which has a piece based on an article in the London Times, it is: “Pawnbrokers to the Rich Boom in Britain” Now if this piece had predated Borro’s current PR push I might have been less skeptical that Borro is meeting a genuine demand for rich people to suddenly use pawnbrokers. But it doesn’t, and there are no other facts in the piece to backup any trend. Note the headline isn’t: “Pawnbroker to the Rich Booms in Britain” it’s plural, as if Borro is the winner in a vast new space. It isn’t just claiming that Borro is booming, but that the rich are also busting.


Its in the journal so it must be true.

But the rich are not going bust in droves. In fact one of the defining trends of the Great Recession is that the gap between rich and poor has widened. If, as the WSJ article says, “borro is sending tow trucks to posh enclaves like Kensington, Chelsea and Fulham to pick up an ever growing list of Porsches and BMW’s”, why have property prices bucked the national UK trend and steadfastly continued to increase there? Sure, the business of pawnbrokers may be thriving, expanding their client base to nip at the heels of the middle classes, but I doubt the last resort lender of the bottom percentage has become a favored recourse for the 1%. In fact the 1% are more likely to be found funding sites like Borro, these days, with nowhere else to put their money.

Here’s what I suspect happened – just my opinion. Borro hasn’t been around long enough to be sending tow trucks to pick up an ever growing list of Porsche’s even if it had struck gold by inverting a thousand year old model of lending. But they’d like to be and that is the story their PR agency is selling. Via the distortions of distance, cultural difference and the persuasiveness of PR flacks, a Wall Street Journalist was duped by a story that if it were true would be emblematic of the bigger picture, of a shit storm of debt default which can only be understood via anecdotes like this. It was a story the writer wanted to believe, it slipped through, no big deal.

But if a single missing ‘s’ in the WSJ makes this story fascinating for an OCD pedant like myself, a single letter typo on their website would be enough to excite Benedict Cumberbatch’s Sherlock.

Borro have a list of things that people have already pawned, from Chateau Lafite claret to a Banksy print and a $40000 painting by Jean Dubuffet. Borro point out that the items “are not accurate representations due to client confidentiality however the loan amounts and artists are precise”.

Now aside from the fact that the confidentiality is convenient for a new business with little loot to show, or that DuBuffet paintings are about $750 per square inch meaning it was on the small size (less than 8 inches square), just like Arthur Daley’s memorably dodgy Levys [sic], DuBuffet isn’t spelled the way Borro did: DeBuffet.


DeBuffet is spelled DuBuffet


No big deal I guess, many sites have typos. Perhaps their picture expert did dictate the details and passed on no emails with what they had sold. But in the world of Occam’s Razor, I’m not buying it. Forgive me if the thought hasn’t crossed my mind that there are no DuBuffets in Borro’s trembler cage vaults, no lines of trucks waiting to haul Porsches out of Chelsea parking spaces and that neither the business of pawnbroking has become upmarket nor have the 1% become the latest victims of austerity.

Even if Borro is just a dressed up pawnbroker why is it intrinsically worse than a bank or other distressed debt sites like Wonga?

Banks and even Wonga don’t want you to go bust, or they don’t make money. Some people won’t pay, and the debts will have to be written off, but most will pay back and the net profit will be the interest minus the rate of default. But if everyone defaults on a pawnbroker they can still make money.

On the face of it, Borro aren’t trying to profit from a default. From their values section: “[We] Realize the highest value for our customers’ assets in the ‘last resort’ event of default, so that their loans are repaid in full and any excess is returned to them”. But think about that. If the loan is at a very high interest rate (I’ve seen figures like 60% bandied around) then the selling of collateral is already yielding a big profit before any money is handed back to the borrower. More importantly, with the hand back, there is no incentive for Borro to sell for a maximum beyond the loan value, and businesses where goodwill is in conflict with incentives are either not good will or not good business.

As I’ve written before, the last Internet boom failed when people though they could wave a magic wand at any offline business and Internetify it, bringing unimagined riches as keyboards clattered like a million cash registers. Borro looks to me to be making the same mistake, trying to take pawnbroking online and make it something that it isn’t from a business perspective and which fails to understand how the web works from a product one.

Why does all this matter?

This matters to me on every level. As an Internet entrepreneur, Borro is the kind of mad business that helped creat the last crash. As a technologist, if that happens again it will destroy faith in one of the things that can help get us out of the current mess via innovation or the desire to make things simpler and better such as Simple (BankSimple). And as a person, Simple in turn reflects the ability of technology to solve rather than perpetuate the money-for-its-own-sake mistake of the debt-fueled, financial arms race that nearly destroyed the global economy.

More than that, by having, what is suspect was harmless PR driven fiction enter the archives of the Wall Street Journal, with the title “Pawnbrokers to the Rich Boom in Britain” and the date “January 4, 2012” this innocuous article could be re-quoted to rewrite history and create the cruel illusion that the 1% were in this together with the 99%.

I Used to Love Trains

Posted by | Uncategorized | No Comments

I used to love trains, in secret I think that I became an architect because I spent so much time building infrastructure for my model railway set, as a kid. Having watched trains crawl along at the speed of slow freight, in my 10 years in the US, I moved back to Europe and was excited by the prospect of regularly using a truly high speed rail network.

Then I actually used it.

A train on France’s TGV from where I live, in Geneva, to Paris takes a mere 3 hours, it is by far the best way to get there. But it costs almost double what a flight does. If I want to use the Swiss network it’s more, and if I want to combine two networks, say by taking the train to Berlin via Basel (which I have done, when flights were delayed by snow) then not only does it take a full day, but the idiosyncratic differences between providers means that the complexity is several orders of magnitude higher than a multi-point plane flight. Ticketing for the TGV is worse, positively Byzantine. There are no print-out tickets, let alone iPhone barcodes, you have to pick up or have a cardboard ticket delivered.

So what’s going on, why is state subsidized, cheap, electric powered transport more expensive than for profit air travel at a time when gas prices are high? This is not just a function of competitive efficiency, although that is part of it, (SNCF went on strike, for example, to prevent freight running on the TGV network), in the UK privatized rail has created prices that are even worse.

The problem is the infrastructure itself. When you fly a plane there is no track required, the infrastructure investment is only at the nodes, and the decentralization of these nodes by low cost airlines in Europe, has driven prices down and broken down the stranglehold of the large airport authorities. On a rail network there is a massive cost for the rails themselves that creates less competition so privatization means monopoly capitalism.

So why isn’t the same true of the road network? That requires massive infrastructure and in France the road network is possibly the most free-market driven infrastructure in the world with private toll freeways operated by companies in places like Dubai. The French road network is ironically far more capitalist than America’s. But it isn’t the nature of the ownership of the infrastructure but the nature of the network itself that determines its character. Railways can only pack a very few number of trains on at one time and there are few exits or stops. Roads have many vehicles like little packets of transport. The road network is like a packet switched one, like the Internet whereas the rail Network is like the legacy, fixed line, monopolistic telephone system.

But at least trains are green, like the nuclear generated electricity that powers the TGV.