What a year this has been. Back in February, when the HAT project ended, and was passed from the 6 Universities to the HAT Foundation, we knew that translating 6 briefing papers, developing the tech and implementing the design principles into an actual ecosystem was going to be a challenge of huge proportions.
2016 was the year we embraced the challenge. 2016 was the move from design to implementation; from concept to capability. From the written word to reality.
2016 was the beginning of a very hard journey.
To begin with, how do we make it real? Much of what the HAT is conceptual. The IDEA was so appealing that the reality must surely disappoint. How does one make control, empowerment and privacy with personal data real, visual and tangible? What does it mean when we tangibilize it? How does it feel like? look like? I felt, often during the year, that we were building something to meet the expectations of so many, that we were obviously doomed to fail. It was like making a movie of a book that everyone had imagined differently in their minds. Here's a chronological account of our thought processes. TL;DR perhaps, but we felt we needed to document it!
February 2016: Building the digital body container, avatar, whatever. I guess we have to start somewhere. Little steps right? We knew we had to keep the idea of the HAT separated from the services that used it. Even if we were building the services as well. So.... practice what we preached. To us, the HAT is a person's digital body. It had its own set of APIs and so could connect with different services, but the HAT was a digital embodiment of the person; owned, controlled and empowering the person, and was therefore the most sacred entity. We wanted a world that respected that. And we wanted a world that built services to respect that. So we built the HAT with containers, so that there are clear legal and technical boundaries. (See tech and security architecture). And technically containerising HATs also made it portable, so that the individual would never be held captive by any HAT provider. We also started to embed the containers in the cloud, using Amazon web services, ensuring, every step of the way, that no one could peek into the HAT, without the individual allowing it. And all other services that used data on the HAT, were all separated from it, and communicated with it through a set of HAT protocols.
March 2016: Indiegogo - were there others that thought like us? We launched Indiegogo campaign to see if there were other like-minded people. And by 16 April, we hit the target of £50k. Great. Now, some 300+ people want HATs and are supporting us and we still have no idea how it looks like. Building continues....Faith is a wonderful thing...
May 2016: Can we see this HAT thing, please? We started building the HAT's first service. - Rumpel, the HAT dashboard. Gus joined the team, together with Mike, who designed the UI. Rumpel finally allowed a user to 'SEE' his HAT and made a HAT tangible for the first time. We can click on something. We can see our data. Yay us....
Rumpel brought in a whole new set of problems. Yes, there were 11 design principles from the research. But so many other questions remained. If it was a dashboard for HAT data, how should it be presented? Would users be happy to see their own data when they see it? How does it look like on the HAT that could be different from how it looked like where the data came from? What was the point of it? Stop shouting 'use case' at us! We're still building the platform! With Rumpel came the invention of new 'terms' that didn't exist on the Internet, but needed to have a name. Like data exchanges. How should it feel like? Should it feel like the way we exchange money? Or should it feel like the way we fill forms to give our information away? At some point we thought about the people who invented modern banking and invented terms like direct debit, standing orders, credit and debits - who named them? Naming became an obsession for us. Naming something that was too alien wouldn't help with adoption. Naming it too close to what exists risk misunderstanding it.
Then came MarketSquare. The principle business model of the HATDeX was to facilitate data exchanges, so MarketSquare was crucial for HATDeX. Even though it had a web front end, many of the middleware services on the HAT ecosystem (e.g. collecting, sorting, allocating metadata) had to be built before any service could be built on the HATs. That meant we had to name and explain again. Data debits describe the way data is exchanged. Data offers for those wanting data, data plugs for data coming into the platform. DeXter was the service that crawled the system to find out which HAT wanted to share what data with whom. HAT2WEB posted stuff from HAT to websites..... argghhhh! We could just hope that everyone can understand what they are. Even in simplifying, we needed to think about how to educate through interactions. We felt the weight of our own aspirations. Like an albatross over the neck....
June 2016: ANOTHER sprint? Many sprints at HAT central. Thanks to those who fed and watered us. Loads of late night discussions. And of course alcohol.
July 2016: Release? And Crashed! We have something half decent but very buggy and released HAT and Rumpel to 136 beta users. Boy was that a disaster. Half didn't understand what the HAT was meant to do, the other half couldn't get it to work. We worked through it. Bug by bug. Functionality by functionality. Created some screen tutorials. Probably alienated quite a few people who wanted something more 'polished'.
We knew the dashboard and the exchange infrastructure was completing soon, and it was time to design the first ervice on the HAT that could show how it would be useful (even though minimally) and would serve as the first point of adoption for users and yet showed what the platform could do that was different from others. Yes, yes... it's use case time.....
Shout this out loud to feel the frustration of the HAT product team: "It's the person's digital body. If you want me to tell you the use case, why don't you tell me what your physical body's use case is. Have you figured out what you're put on this earth for?"
Our CEO very calm response to the product team....." we do need just ONE......."
Point taken. Humph. The thing about inventing something is that there really aren't many useful guidelines on how to create something new. We felt like artists with a colour palette and a blank canvas. Time for long bike rides.
August 2016: Notables born. At last we had ONE. It wasn't perfect, but it was something. It meant we took a very different turn with personal data. We came to the conclusion that personal data as we knew it was the most boring thing in the world and went towards personal data that was more meaningful. Words. Numbers. Images. And we started with words. Notables was born. With Notables, we felt it ticked the boxes to show off the platform. There was meaning in the personal data (you see your own words). There was sharing (posting on social media). And there could be analysis (with sentiment bots for mental health for example) for both the user to benefit and for sharing with others. That meant a singular use case for the HAT to go into the public domain that could demonstrate what a HAT-enabled future could be about. It also meant people might just use it.
September 2016: Getting interest.... We had conversations with potential HAT providers. An insurance company. A bank. Things were moving. It's amazing how things move when you can touch and feel something.
November 2016: Wider release... Marios joins us to build the iOS app Rumpel Lite. We released a proper working application on web and on iOS for HAT and Rumpel w/Notables. We further fine tuned our message. And on the data plugs side, we were bringing in Facebook, Calendar, Location, Twitter, LinkedIn and others....
All through the year, we battled with how we should describe the HAT in simple ways, without dumbing it down too much. Personal data wallet, personal data platform, personal data server,..... so many terms came and went... until someone said - why not just call it a HAT :D. In the end, we stuck with personal data store to describe what it is, but committed to calling it a HAT and to make it our mission to get everyone to call it a HAT. One. full. circle.
December 2016: Ready now....The platform is finally ready to start scaling in the new year, and we move into the next phase of getting market traction and next round of Angel funding. The HATs in the backend are being optimized for resource usage and the system stress tested. Based on conversations, 2-3 HAT providers will come online next year offering HATs to their customers and HATs will begin the scaling up.
The year was more art than science. As a multi-sided platform, we knew we had to have a proposition for firms and for individuals that was beyond just privacy and control. We now have it for both. Firms here. Individuals here. We knew we had to mimic services on the Internet and yet create something that didn't yet exist. We had to do it visually and with interactions. It had to achieve an outcome where the user feels familiar interacting with it, but yet feels he is doing something completely new, and yet useful. If the interaction was too new, it would deter the user from clicking on it. If it was too familiar, the user may not feel any different from another Internet service that used and abused their data. We called the build strategy a combination of mimicry and scaffolding. We mimic the old, to scaffold the user into the new.
After a year of coding, designing, eating, drinking, sprinting and fixing bugs; a year of long bike rides, cucumber water, train travels, HAT central cooking, barbecues, pubs, wet socks (the team's term for a 'not-quite-there' product) and a lot of arguing, we've built it. With the platform now built, we know we have to incentivise the creation of day-to-day use of services on the Internet - calendar, social media, emails, chats, browsing - so that privacy seeking individuals have a choice, a choice of bringing control back to themselves while still getting the same services they now have on the Internet. A platform that, as Internet users, we never had (and therefore it was easy to exploit us, data wise). We are optimistic because advances in tech are on our side. Our community can build services faster and better because HAT centric Internet services could actually mean better Internet services than the current ones. With the HAT technical infrastructure now built, we can look forward to services that not only consolidate our own data for messaging, for calendars or for browsing, we can bring in more data - IoT data, finance, health and other data that the Internet currently doesn't have. We can build AI bots on our data and other cool services. We do not need to rely on apps built on GAFA for services. There is now a choice. For future app builders and future Internet Services.
We have built the capability we talked about in the last 3 years.
The HAT is both a tech and a economic/business model innovation. The HAT is now beyond slide ware, books, papers, and power points. It now exists. A full iOS app and web version will be released in January 2017 and marketing will start. 2017 will finally be the time for HAT services. And community. We would like to involve the community of app makers, startups, digital natives, and liberal-minded individuals who believe in personal data empowerment, and who believe in an Internet that is very different from what it is today. A world where many firms would build services on HATs because there are more revenues in data exchanges than selling data for advertising. Most of all, we want to build a community of people who wanted a change from this data exploitation regime that the Internet economy has become. We are clearly a David amongst Goliaths, but we can only try.
2017 would probably tell us if we can succeed. Happy New Year from the HATDeX team.