Jul 26, 2018 at 16:19
Jul 26, 2018 at 16:19 UTC
Up Close and Personal with Dimitri De Jonghe: What Human and Machine Interaction Could Achieve
Jatin Madhra, the founder of Bient Technologies, was in conversation with Dimitri De Jonghe, Head of Research at Ocean Protocol and BigChainDB and Co-Founder at Ocean Protocol.
Founded back in 2017, Ocean Protocol is an ecosystem that has been built for sharing data and associated services. It aims to unlock data, particularly for AI. It uses blockchain technology to facilitate data sharing and selling in a safe, secure and transparent manner.
Dimitri also heads the blockchain development department at BigChainDB, which serves as a database for blockchain. It allows developers and enterprises to deploy blockchain proof-of-concepts, platforms and applications with a blockchain database, supporting a wide range of industries and use cases. Following is the transcript of the interview:
At CogX2018, you were quoted as saying, “It’s interesting to think what happens when machines can own resources and decide what to do with them.” Kindly elaborate what follows if this happens.
AI and Machines are intelligent technologies that have been working in congruence with each other for some time now. We went one step further and allowed it to transmit values. So this means that there is a possibility that it could be transformed into something that doesn’t have a guardian anymore and still owns resources and generates value in the ecosystem and connects values from other ecosystems. So in essence, not having a human guardian but still being an economical agent or maybe a legal agent in a system that’s not human is a very interesting thing to have a look at. So suddenly everybody is talking about self-driving cars, but the first thing that happened is self-owned cars, with some other kind of logistical infrastructure, or what’s even more interesting is to look at natural resources. Some stuff just doesn’t belong to humans, because it belongs to the world, humanity, to the commons, the public. For example in New Zealand, some rivers have a legal status. And now I have a project like ‘Terra0’ which works on nature. And with recent developments in Malta, we can witness humans interacting with non-humans on different levels. So we know information sharing, we know how to read out of sensors and all of these things, which gradually push us to become more autonomous. They need access to a system that humans cannot override so that they can choose how to spend their resources and maybe even hire humans to do something for them. Let’s take the example of a self-owned forest. So a self-owned forest basically wants to become self-sustainable; have enough trees inside of it so that it can keep on growing or at least keep a capacity of trees. So a human can come in, it can take a tree, but it has to pay the forest for that. What can be done here is that maybe they can hire a set of drones, or a set of wood checkers, to keep the forest mapped out and all of these things. I think also maybe monuments. So I mean yeah, that’s kind of the things that are happening and basically we are facing a world where machine human interaction is about resource share and not just about information exchange.
We’re well aware of your extensive experience as a co-founder of Ascribe, Application Director at BigChainDB, Researcher with Ocean Protocol, and of course co-chair of the Interledger Community. But Sydelity comes out as the most striking of all. Please elaborate, in layman’s terms, what exactly Sydelity does and how it started?
Sydelity is an old venture, it was my first startup. You could say that it is in sort of a sleep mode but it’s actually a spin-off of my research at the university for my PhD where we are always trying to find efficient mechanisms to simulate very complex non-linear dynamical systems. In this case we were looking at microelectronics. Imagine an intel processor which has a lot of transistors on them, more than a billion transistors. So all active components that have to interplay, and during the design phase you want to predict how these processes fit in. If you want to work on the functionality requirements, you have to basically simulate those during the design phase. Let’s say if you want to add more transistors or change the floor plan, then you just want to make sure that you don’t make any goof ups. So what Sidelity does was a bit like what jpeg does to high resolution images. Sydelity does the same to complex floor plan systems and dynamic systems. It compresses them without compromising too much on the details or the functional behavior. So, it’s basically applied mathematics for complex dynamical systems. So if we look at Token economies as a complex non-linear dynamical system with user behavior, imagine that we are creating a token incentive schema, and then have all types of behavioral agents attached to it. Now the question is how will this incentive schema turn out? Will this become a self-sustaining network? Will it die out? How can we play around with the objectives in Token ecosystems. So we need some simulators there to maybe bring applied mathematics back into the game. So yeah, that’s in a nutshell.
So let’s move to Blockchain. Now that we have your insights in brief on ML and AI. Bill Gates and other such tech giants have claimed that Bitcoin is a bubble. Why do you think it’s not?
There’s a bubble aspect to it but at the core it’s not a bubble. At the core, it’s basically one of the strongest supercomputers that humans have ever created and especially it’s an incentive machine. The incentive machine basically means that it makes people do stuff. It makes people buy mining infrastructure. It also creates an open market for mining infrastructure to add security to the network. That’s the objective function of bitcoin. So it’s a precedent. It’s not a bubble because the utility below is multifold and it’s super highly secure. It avoids double spending and it’s also built inside a community, sub-culture and of course the aspect of be your own bank is novel. So it’s a precedent, a paradigm shift so that people can collaborate on incentives.
Great! I’ll move on to Ocean Protocol. Ocean protocol recently partnered with Agorai, an integrated marketplace for AI. How do you exactly plan to grow, or to increase the adoption of Ocean Protocol apart from B2B partnerships?
So, these B2B partnerships, network to network or protocol to protocol partnerships are not really businesses. They are just trying to create more flexible systems. That’s a big part of it because if you want to create something like a public intelligence system to solve all problems of humanity then we need to have a lot of moving parts in the system and we have to bridge and collaborate with a lot of other networks. What we feel we should focus towards is a very well designed incentive schema for people to have early mover advantage, for people to stabilize the system, to adopt, collaborate and communicate. Especially if you need to look at it from a functional point of view then we are trying to create something we call tribes. Tribes gather themselves around a problem space. So let’s say we could have a mobility tribe, trying to solve the problem of autonomous driving. Or making mobility more than just about transporting humans, but transporting energy and stop congestion etc. So, all these tribes focus towards fine tune this network, to meet the local needs of any party. So, it could be a group of data scientists, or a group of humanists with their own governance. So, we just give them to tools to determine how they share their data, or what algorithms they can access, etc.
So a lot of blockchain development conversations focus on developing a user centric application with blockchain technology at the core. But what I personally believe is that for about another 2-3 years, blockchain will be limited to being a backend technology and will take time to dissolve into that position. So, you being at the epicenter of creation, when do you think normal user-centric applications can be built on blockchain technology? For normal people who use Spotify or Uber or various food apps on their phone, when will they have access to blockchain based applications in their daily lives?
I think the need of the hour is to build developer centric platforms. The reason internet went big is because it has a standard language TCP/IP. We are still trying to figure out what’s the standard language for inter blockchain communications and smart contracts, so we don’t have enough tool now to impart to the developers. Once we do that, then you’ll see architects, functional analysts coming in and starting to go from this place small contracts to creating compound contracts and basically making the blockchain layer seamless. And then you can think about creating user centric applications.
I’ll move on to the last question for the evening. Is it safe to anticipate BigChainDB and Ocean Protocol working in an integrated manner in near future?
So this is exactly what we are doing at the moment. With Ocean protocol, we have a lot of assets to be published, lots of metadata. BigChainDB is an interesting tool to store and decentralize that metadata. For us, at the core of our discovery, there you will find BigChainDB, at the discovery core of our function. We have a vision we want to realize and we wish to contribute value to the ecosystem by leveraging all the technologies we own.
The most important takeaway from this interview is the way machine and human interaction could be enhanced and channelized in novel directions with the use of innovative technologies that we have at our disposal.