Sergei Lonshakov, the developer of Airalab, has presented Aira, which is an acronym for Autonomous Intelligent Robot Agent. The project implements economic interaction standards of human to robot and robot to robot with a unified obligation contract.
It is conceived as a foundation of a new ‘robotic economy,’ which, according to Lonshakov, is to become the driving force of the future’s global economy.
ForkLog talked with Sergei to find out more about the project and the robotic economy to come.
ForkLog: Tell us about your new project.
Sergei Lonshakov: We’ve been working on the borderline of robotics and digital economy for a while now. There was an initial understanding that we wanted to build direct economic relations between humans and robots, so we’ve had numerous experiments in this area. This year, we realized our research could be of more use than R&D works.
These days, it’s not machines living in the human world, but humans live in the world of machines. There are more machines than us. They’re a part of our economic relations, and we constantly use machines to solve our problems, however, communicating with robots remains an enormous problem for robotics. Aira implements human to robot and robot to robot economic models with an obligation contract. Aira allows connecting numerous different robots to an obligations market in Ethereum to directly sell sensor data, ensure drone delivery, and order customized products from robotic enterprises.
We take economic soundness of such a market structure into account in terms of micro and macroeconomic factors, which is very important for any market structure’s viability.
FL: What about regulation? Are there any government solutions in this regard?
S. L.: In our communications with corporations, I see them discussing the issues of robot regulation. In most countries, there is an active dialog suggesting that drones have to be operated by the nation state. All corporations are now very interested in deploying cloud computation solutions which would incorporate decision-making centers for robot activity. In other words, when you approach a vending machine, you interact with a remote decision-making center that gives an order to the robot next to you as it were a sensor.
FL: Should the global robotization project and the internal economy of robots be within the government regulation framework?
S. L.: On the contrary, a central authority is a hazard. We see a consolidating trend where robot operation is being centralized. It is downright wrong because it would result in total control of every human being, i.e. a minority would rule majority. Thanks to the centralization of robot operation we might lose the social equality because only certain people would own such an important attribute of life as robotics.
FL: Nation states aren’t quite happy about giving away the global control of society and economy, you know.
S. L.: They sure aren’t. The most important thing here is that there’s an alternative solution in terms of government control that makes everyone happy: it’s market relations. A government is a part of the system, not some sort of an add-in. We’re not trying to invent something new. We’ve seen totalitarian regimes where the government put itself above the society and replaced direct economic relations with control. However, dictatorship is inefficient, it’s very bad at distributing resources and tasks within a society. We offer to use the existing experience as an alternative to the centralized control: a free market. We have to build market relations and develop a robotic economy on top of them.
FL: Robotic economy may bring about both positive and negative changes. Aside from obvious massive loss of jobs, what are the other problems of such a structure?
S. L.: We have to understand that robotization is inevitable and will result in serious rearrangement of society in any case. That goes well beyond loss of jobs. However, what we’re saying that, judging from the experience, it’s better to introduce machines in the society economically, with their own market. It would be much safer than total centralization of control.
The first step is to establish a market for sensor data sales. As an experiment, we’ve released a Proof-of-R&D. We’re about to deploy a few sensors: a Geiger sensor at a St. Petersburg-based university and a weather station at Togliatti. Those sensors would directly sell their data for our internal tokens. Any person connected to the Ethereum network may visit our DApp and purchase said data.
We believe that it’s the sensors that have to be installed in the first place. When there are sensors of weather, noise, pollution, traffic jams, airspace availability and so on, it is a basic platform for deployment of ground, maritime and air transportation systems. If we have a full picture of what’s going on in the world, and one may easily buy said data, an automatic van may purchase sensor data to deliver goods with no trouble.
FL: Alright, sensor revolution is the first step. What’s next?
S. L.: We’re making three releases now: first, a sensor data market; second, connecting a market for safety of ground, air and maritime transport; third, there has to be a market for selling products of fully automated factories. This is the last step that activates a full-fledged supply chain that we envision: you order a product, a factory starts assembling it, and then automatically sends it to the address you’ve specified.
FL: Is there a technological base for the project? How would a multitude of robots interact?
S. L.: We want to create a robotic economy, but now it’s mostly about some basic interaction. What we do with Ethereum and ROS (Robot Operating System) is a framework that may underlie any robot. There’s a big academic community around the ROS. There are industrial solutions that are potentially ready to use the ROS for creating a working factory. By supporting the ROS we automatically get factories that can work with the market in Ethereum. We’ll allow any ROS-compatible robot to install Aira as an assistant. It will be automatically capable of selling its obligation to perform some work. Additionally, there is a prediction mechanism in the foundation of robotic economy.
FL: What will be your next steps? Are you going to run an ICO?
S. L.: We think of running an experimental sale of Air tokens. Tokens usually reflect market capital. For instance, Golem tokens reflect the potential value of a distributed computation network market they will create after deploying the software. That’s called a market of ideas. People invest in an ICO as an idea and get a market of products out of it. We see that the process has a futures component in economic terms. The thing is that prediction markets are a more precise instrument of assessing potential of anything than voting or polling. When people vote with their money, it’s more likely to be a well-weighted thing.
We’ll make a multi-vector market with a prediction mechanism for a limited capital. We believe this would allow us to build a network of self-adjusting robots.
Imagine there’s a robotic factory producing flower pots. If the factory monitors the investor assessment of the next period, it may rearrange itself to produce more products, or differentiate the production facilities if people have relevant expectations. And then it might start producing clay figurines of happy elephants. That is the experiment. We see that with Aira and our approach towards its implementation such a network of fully automatic economy may exist and react to changes in human needs.
FL: Are there any basic rules for the robotic economy?
S. L.: In our experience, we’ve found three basic rules to make the economy secure and dependent on human economy. First, a robot can’t have ownership rights. As long as robots work solely on the basis of contractual obligations, humans can run robotic economy. Second, robotic economy has to have internal capital. It’s the internal capital that can maintain the market and reflect the value of robots’ direct involvement in human lives. Third, robotic economy may exist only within digital economy that ensures functioning of internal capital and fulfillment of contractual obligations.
FL: If the internal robotic economy uses Ethereum, will there be an internal token belonging to the robot’s owner?
S. L.: The robot’s owner will get rewards for the services rendered by the robot in those internal tokens. Generally, we want to sell tokens to early investors to generate a market assessment of the capital. With such assessment, market players are likely to join. If you have a robot and you’re willing to sell its services or data, you have to have an easy way to do it in order to spend as little time as possible while knowing you’ll get something in return.
FL: Roughly speaking, if you have a robot you may connect it to the platform where contracts for its utilization will get automatically concluded, right?
S. L.: Precisely. You have to download Aira image to your computer and set the price. Then the robot will automatically go to the market and place the offer. With the DApp we’re going to make buyers can connect to the market and choose the offers they find interesting. Once a person buys some data the robot will know the contract payment has arrived, then it will perform the task and report as to the task in the contract. We use an IPFS network for that purpose: the robot performs the task, puts the results in the IPFS, and as the IPFS file’s hash is immutable, it publishes the obligation contract. And only once the hash is published the beneficiary will get the money, which is Air tokens.
FL: Why do you think investors would find Air tokens interesting?
S. L.: Aira image will be directly linked to particular contract addresses in Ethereum, to a particular market and a particular token, which will maintain the token’s circulation. When someone invests in any token, they see a rateable value. Then market players come along, and the buyers need the token. Thus the demand forms. We assume that the robotics market is expanding, and there are lots of people around the world willing to connect to decentralized networks and sell their robots’ services. However, there has to be a market considering the main macroeconomic challenges. Those challenges are as follows:
- How do we ensure transfer of rights if there are no people involved in production and delivery?
- How can a purely automatic factory understand ever-changing human needs?
- How can robots of two different corporations directly interact?
- How to tax robot work, and what is a separate robotic unit?
Our model answers all those questions, so I’m sure Air tokens will find their demand, and we’ll build a safe robotic economy.