DFINITY: A Hack-Proof Platform. Talk with Dominic WILLIAMS, Part 3 of 3

This is Part 3 of my conversation with Dominic Williams. Dom and I sat down to talk about how DFINITY got started, how he came up with the name, and most importantly: where we’re heading to. Join our conversation about both the milestones that we’ve reached and the ones that lay in front of us!

Video Transcript:

Cédric: You’ve touched on two very interesting points there’s as well that came up in a recent Q&A from the community – one is interoperability and the second is hack-proof and cybersecurity. It’ll be super interesting, I think, for the people watching if you could say a few words about each of those.

The Internet Computer

Dominic: So the Internet computer looks like a computer, it smells like a computer, you can install software on the Internet computer, you can execute that software, it’ll store data, and so on and so forth. But it’s not really a computer – it’s a kind of a virtual computer created by this massive network; full nodes that run the DFINITY protocol. What this means is that this Internet computer is deliberately, in many respects, very simple. That’s what you can do – you can install the software, you can run the software and again it can store data on. Then with a traditional computer you can open up a terminal window and poke around, and try and escalate your privileges, and maybe, modify the contents, the password file or something.

That’s just not possible on DFINITY because all that stuff doesn’t exist. I mean it’s a virtual computer. Really it’s a figment of the protocol execution very different to a traditional physical computer with an operating system or even a virtual computer built on such a thing like an EC2 instance. So, the platform is hack-proof.

Hack-Proof Platform

Dominic: What does that mean for the people building on top? Well, at least you don’t have to worry about the underlying platform being hacked. You do have to be cautious when you build a software system on top if you create a function that anybody can call. Anybody can call it. You also have to make sure that your users, who have some kind of public key identity, don’t lose their private keys because that would enable someone to impersonate them an access.

Notwithstanding all of that, obviously, so long as your software doesn’t have bugs and so long as your users don’t lose their private keys and so on. Your systems will also be hack-proof. You don’t have to worry about firewalls or anything like that; just the computer itself is inherently hack-proof or so that. Obviously, it moves the game forward in a major way, in my view. And especially when you can combine that with technologies that preserve the privacy of user data, you begin to see how this could really be revolutionary in a number of different fields.

Privacy of User Data

Dominic: And we’re all aware of horrendous hacks that have occurred. The Office of Personnel Management, for example, losing pretty much all of these sensitive details of the people, who have applied for jobs in the US government that have some security sensitivity. That’s just exactly facts and it just goes on and on forever with all these hacks.

I think we’re seeing a sort of collapse of the traditional IT stack with all its complexity in the sense that it just can’t provide people with the security and the resistance to hacks that are necessary; and it can’t look after sensitive private data. So I think anyone wanting to build a new software system will definitely find all of these features – security, availability, interoperability and privacy extremely attractive. So in that sense the Internet computer is just trying to reinvent how we build software and host software.

DFINITY Autonomous Software

Dominic:  But there’s also another interesting dimension to it and that is now we also enable the creation of completely new kinds of software. The most important is what we call autonomous software. Autonomous software is a system that exists independently of its creator or any other “owner.” Typically an autonomous software system will be built with a governance system, which is controlled by tokens and through this governance system the autonomous software can update itself. So in that sense we’re applying the open-source software development model to a running Internet service.

So at the moment open-source is only good for updating some software package. There are a number of reasons we think this will eventually be huge – one is that you can provide guarantees about user data security, an Internet dating service that kept all of the chat database in a crypto lock box and we have a way of doing that that allows algorithms to be run on the chat database. So, for example, looking for matches between the users, who maybe have common interests, by analyzing what they’ve said, while ensuring that the contents of that chat database can never be extracted. They just always stay in the crypto lock box.

Open Source Business Model

Dominic: So an Internet dating app built on DFINITY using the autonomous software model – the open source business model, if you like – can guarantee to its users, “Look, don’t worry. You’re going to be on this for a few years, and you even may even settle down and get married, whatever it is. The contents of this Internet – this embarrassing chats you have – is never going to come out. So that’s just a very simple example of how autonomous software can provide some interesting guarantees to end users.

But actually there’s another really important piece. Today’s the Internet has become very different to the Internet of old, the other networks that weren’t so successful. So back in the day – in the 1990s – there was a lot of debate about whether something like the Internet would succeed or private networks like AOL or CompuServe. And in the United Kingdom, British Telecom, which was a sort of monopolistic telecom provider was trying to persuade the government that they should be allowed to create this thing, called the information superhighway and connect everyone in the UK up to their own private network and there’s going much. The Internet is dangerous; you can’t control us on it, we’re going to create as wonderful walled garden and the Catholic curate all the  content. Thankfully it didn’t come to pass.

Two Reasons for the Internet Success

Dominic: Now the reason the Internet succeeded is… There are two reasons. Because it was an open and decentralized network – an open and decentralized network without
any central point of control – all of these ISPs could build out the network. So it grew very fast. No one needed permission. In the UK back in the 1990s there were hundreds of individualized pieces, who built out the network and provided access to people. The Internet was able to grow very quickly and provide access to lots of different people through enlisting the help of lots different parties. That was the first thing.

The second thing was that the Internet was an open network. So that meant if you’re an entrepreneur, you could create a new service and connect it to the Internet at very low entry cost. If you were successful, you didn’t have to worry about a competitor degrading your access to the Internet or degrading your service in some way. In contrast, if you wanted to build a service on AOL, number one – you’d have to persuade AOL that your service was going to be good enough that they should list it, and that’s going back to the dating app example. In the case where AOL then decided that they wanted to run their own dating app because the Internet dating is very profitable, guess what? Your dating app would be right at the bottom of AOL’s homepage all of a sudden. So it wouldn’t be a level playing field.  The Internet, by contrast, didn’t take anyone’s permission to create a new service. Once you’ve created that service, you’re on a level playing field – nobody could degrade your access.

Monopolistic Internet Giants

Dominic: What’s happened there recently with the Internet is that it has become very monopolistic. There’s been a lot of consolidation around these monopolistic Internet giants, whose services have become platforms. You need really now to use these platforms to gain access to users, user data, functionality, markets, and all kinds of things. You’re not really building directly on the Internet anymore, and you need to build against these platforms that are run by this monopolies.

Platform Risk

Dominic: Now this has led to something called platform risk and this has been killing investment into new Internet services. So I’ll give you a really simple example of platform risk – Zegna, a very successful games company that has created FarmVille, Bruce Tangermills had 200 million users. It was a social game built on top of Facebook. Their IPOed worth 10 billion dollars or something crazy. Then, unfortunately, one day and it’s Mark Zuckerberg or someone else from Facebook said, “We don’t like the way that these mechanisms, these games are able to use to promote themselves, and maybe we’d like to push some of our own social games in, and all this kind of stuff. They just changed the mechanism. So overnight that was pretty much the end of Zegna and it’s never recovered. That’s an example of platform risk. I actually personally knew a company because some guys from the sort of Cassandra distributed database I was involved with and it created like a communications map of an organization both internally and externally.

Anyways, the point was in this communications map, the nodes were people and if you hovered your mouse over one of these nodes, the profile of that person would pop up. Unfortunately, of course, that profile data came from LinkedIn. So this company was doing great – amazing team, great vision, the work became a unicorn, we’re doing very well indeed. Then, one day in 2015, LinkedIn decided that the only people that were going to be able to access the LinkedIn profile data, were going to be Microsoft and Salesforce. Pretty much overnight this amazing company had the rug pulled from under it and they ended up having to sell themselves for less than a quarter their valuation to Salesforce. One of the two monopolies I should point out that still had access to that API.

API – Application Programming Interface

Dominic: Now if you’re a VC and you’re looking for new companies to invest in today that want to build Internet services, typically they’re going to need to use the data of one of these pre-existing monopolies, their functionality or something. You’re a VC, I can see you’ve got a great pitch and this makes absolute sense but what happens if one of these platforms do a Facebook or a LinkedIn. The first thing is we think that with the Internet computer, people can do build things like an open source LinkedIn using autonomous software. The open source LinkedIn built using autonomous software will be able to provide guarantees about the APIs that give other entrepreneurs the access to the profile data.

This is going to do two things – firstly, it’s going to mean that other entrepreneurs now have the security they need to start extending the system and building on top of that API; and secondly, it’s going to give investors the confidence to invest in the ecosystem. So what I call mutualized network affects – the open source LinkedIn will benefit because of people building on top and people are building themselves on top. So this kind of mutualized network effects, I think, are going to drive for the open Internet ecosystem and in the end. I think, it’ll fly past the monopolistic ecosystem and people will remember a lot of the Internet giants of today in the same way as we remember AOL and CompuServe.


Cédric: I think speaking about these platform risks is a great place to loop back to something that we wanted to talk about earlier – interoperability. What’s your vision for how different chains are going to coexist and how are they are going to interoperate with each other?

Dominic: So I’m a bit of a heretic. I mean most of my views are heretical and watch the industry, but I tend to believe that in the different categories, a leader will emerge and the world doesn’t want lots of leaders. That’s why, by the way, Internet monopolies exist. Once you’re on Facebook, you don’t need another social networks, everyone else is on Facebook. It’s easier to interact with other people on Facebook.

Cédric: They’ve grown so big and captured such a big part of them.

Dominic: Totally.  So you get a phase shift in the development of these ecosystems. Originally, there were lots of social networks; originally, there were lots of search engines, but at some point various dynamics kicked in that drive consolidation.

Interoperability of Other Services with DFINITY

Cédric: So you think there’s going to be one big player that captures a big part of the market but, at the same time like, if we compare it to social media – again there’s Facebook which is your online profile, but there are various like niche players like Instagram for pictures, or Youtube for videos – let’s say the same thing happens in the blockchain or decentralized space. There’s going to be these niche players for certain types of applications or smart contracts. What is your vision for how these services are going to interoperate with DFINITY and connect to DFINITY?

Dominic: Let’s say we arrive in a world where you have a few category leaders. I mean Polkadot and Cosmos are working hard to address the interoperability needs that will exist. I also think that maybe yet unforeseen mechanisms that will allow these different networks to talk to each other.

DFINITY Ecosystem Venture Fund

Cédric: Because all the infrastructure that we’re seeing in a space is just being developed. Now there’s nothing that has been established for years and years . What is one application or one type of application and that you are excited about that should be launched on the DFINITY Internet computer?

Dominic: There are so many things you can build on the Internet computer. What we really want people to do is start reinventing the Internet in open-source, autonomous form. I mentioned the Internet dating app earlier and how you can provide users with amazing guarantees about their embarrassing chats never leaking out into the public and guarantees about how their private data will be used. But there are so many cool things you can do, we want to see the creation of Dmail like DFINITY mail, we want to see the creation of open source Salesforce – anything that exists at the moment can often be reinvented as an open source business, where you have an in-built governance system and updates the software that’s controlled by token. That’s why we have the DEF – the DFINITY ecosystem venture fund and there are more venture funds in the works to fund people, who want to build on top of the Internet computer.  I also want to say, of course, municipal councils reinventing their systems for collecting parking ticket fines, things like that.


Cédric: Cool. What an exciting time to be alive! Dom, I know, your schedule is very busy. Thanks a lot for taking the time and coming here, answering questions.  For everyone that’s still watching: if you’re excited about what we are building, check out dfinity.org/jobs. There are lots of interesting roles and positions that we’re hiring for right now. Also if you’ve liked this video, please subscribe below and if you have questions for Dom or anyone else on the team, leave them in the comments.  I’ll make sure to mention them in one of your upcoming episodes.

With that, thank you guys and we’ll talk soon.