Understanding The Gold Rush of Scalable and Validated Data powered by Blockchain and Decentralized AI
Trustless through decentralization
Recently, there is a worldwide huge, i mean HUGE hype for the Blockchain technology, but how does this technology create value and for what use cases does it make sense?
If we ask this question to a huge fan of Satoshi Nakamoto, he’ll reply:
For the first time in human history, the Blockchain technology, allows to transact values from A to B, without having to trust institutions but trusting a democratic network, which is self-controlled through open-source algorithms.
Which makes perfect sense and is proven by its market capitalization of ~$200 billion, sorry $250 Billion, OPS now $150 Billion. BTW, regardless price fluctuations, the Market Cap is huge, but, the purpose of this article is about what’s next!
To understand the exponential growth in terms of values that this technology is making, we don’t have to focus our attention on the technological factor of the blockchain itself, but on the values added by the decentralization.
Decentralization means that every information is validated not by trusting to an entity, but through open source algorithms. Bitcoin for example is validating transactions with the “Proof Of Work”, introducing the concept of transactions blocks replicated and shared by the entire network, without the possibility of editing or deleting the ledger.
The challenge to validate data
The Blockchain technology makes sense for decentralize and validate informations thanks to its “Trustless” core. The main use case in the next years will be to move from decentralizing financial data [transactions] to decentralizing and validate the most valuable asset of this century, data.
Today, even if the hype is huge, the technology is in “Early Stage.” During this technological wave, Blockchain technology makes sense for reinventing what the centralized technology has not disrupted yet, everything that has to be redesigned not only with frictionless ecosystems, big data or AI but with universally validated pieces of information, without the possibility of human manipulation.
In the Blockchain market, every centralized approach that we use today are not useful; this is why next unicorns of this technology will be driven by reinventing how we store values without intermediaries. This new technological era is called “Internet 3.0” or “Internet of values.”
To decentralize data, unlike financial transactions, it is not only necessary to store every piece of information, but also to validate how we track and execute data. For example, when Vitalik Buterin started to build Ethereum, the challenge, was to develop a new Blockchain designed like an operating system.
On the Ethereum Network, like every OS, developers can run applications; each Application has to be decentralized into the Blockchain through Smart-Contracts where to execute functions every transaction have to call the entire Smart-Contract, the request, who made the call and the result.
With Blockchain technology currently it is impossible to create a high scalable application and at the same time obtain the added value of decentralization, due to several technological limits, the most important are:
Transactions per second: Bitcoin on average can sustain 3–7 transactions/s, Ethereum can support 12–20 transactions/s. To make a comparison a visa circuit needs 1,200 transactions/s and an application like Facebook more than 1,000,000 transactions/s
Transaction costs: A Bitcoin transaction costs roughly $ 2/7, an Ethereum transaction without running a smart contract costs about $ 0.3 / 0.8, these costs are unsustainable for an application that needs millions of transactions per second.
Network congestion: A fundamental problem is that all applications run on the same network, this leads to periods of congestion where transactions can increase tenfold as costs or even fail… Do you remember Crypto-Kitties?
Specific cases of use: Ethereum was designed to become a decentralized operating system for applications with general purposes, without focusing on particular functions, something like the state-of-the-art network in terms of decentralization and security, excluding a lot of use cases.
The first generation of Decentralized Applications
A lot of excellent teams are developing technologies to solve these problems and make possible the development of complex decentralized applications in the near future. The challenge to solve these problems is to scale the technology to the mass adoption and at the same time maintaining the “Trustless” standard.
The most adopted solution in last years is the DAG (Directed Acyclic Graph), like the Blockchain technology, DAG is a distributed ledger technology (DLT). In DAG networks, users that make transactions also manage the consensus, usually based on a quorum to reach 51% of participants.
This technology, even if infinitely scalable, is dangerous, because it ’s manipulable, and the consensus is based the trust in the participants. What make Blockchain “Trustless”, is the unique value of validating information through algorithms, with anonymous validators and a safety circuit based on economic disincentives to cheating the network.
Non-Blockchain DLT systems, such as DAGs, do not have validation through an algorithm, the problem is that in this early stage of Blockchain technology, people and organizations have a lot of friction to understand the Trustless Concept.
This misunderstanding is becoming more confusing, because a lot of companies, is speculating, driven by “the blockchain trend.” In fact, a lot of Companies are selling non-Trustless solutions, like Private Blockchains (a confusing name for a shared database) or some DLT solutions, wrongly as Trustless Blockchains.
This trend is dangerous because if an organization start trusting something manipulable, the result could be a catastrophe in the long run.
Solutions like Raiden Network for Ethereum and Lighting Network for Bitcoin were born to solve network limits related to transactions per second and transactions costs without losing the security of a Blockchain network. Basically, these solutions are hybrid from Blockchain transactions and a DLT private network.
Protocols like Raiden, allows users to open a private distributed connection, where they could validate on the blockchain only the first and last transaction. The risks of fraud compared to a DAG are limited, but every middle transaction is not validated.
This solution can be useful for micro-payments and to scale manipulable and dynamic data.
A non-hybrid solution, to solve the problem of scalability, promoted by Ethereum is Plasma, a protocol for decentralizing federated blockchains. Federated Blockchains, used mainly in internal company protocols, are infinitely scalable but at the same time centralized, since they rely on a limited number of selected and non-anonymous validators.
Plasma offers developers the ability to create and customize their private Blockchain, through Smart Contract, with or without issuing a new Cryptocurrency and with or without federated validators. The Ethereum’s Public Blockchain controls the Trustless in this solution, as parallelism: Ethereum is the “Supreme Court” and a private blockchain a “State.”
From a vertical Blockchain model to the Interoperability
The demand for the development of the first generation of decentralized applications is growing exponentially, this started a race to develop a scalable and Trustless technology.
In the last two years, a lot of innovative projects start to build technologies for testing new methods of Trustless validation, such as Casper, NEO or EOS, and new Blockchain technologies focused on specific use cases, such as IOT, Big Data, Hosting, etc.
This race has generated the concept of the third generation of Blockchain technology, focussed on the idea of “Interoperability.”
The Interoperability concept, developed by projects such as AION, Polkadot and ICON, opens a new smart way to solve tech limits in Blockchain technology, introducing the idea of unstoppable applications that spoke through multiple Blockchains.
This solution is opening the way to the first real generation of complex decentralized applications, which can use the most performing blockchain technology for each function based on every kind of uses case. At the same time, this new approach can mitigate risks like being at the mercy of the congestion of a network or destroy an entire business due to a network defect.
The Interoperability infrastructure allows developers to create Decentralized Applications in a new smart way that I like to call “Blockchain as a Microservice.”
The “BAAM” concept is similar to the way in which we create modern centralized applications with microservices based on dedicated and replaceable APIs for each function. In Dapp we will use many dedicated and replaceable Blockchains focused on specific needs to perform each task. This type of infrastructure will create an exponential research on the evolution of this technology, especially in new Blockchains focused on specific use cases, without loosing the Trustless validation.
Artificial intelligence, why without Blockchain is limited?
In the Artificial Intelligence industry, another super-hyped tech field, the challenge is different but at the same time vast and complex. In this context the quantity and the quality of data are crucial, trusted on AI algorithms with manipulable data is dangerous for every kind of use cases.
The best results in the AI field are in closed and well-defined ecosystems, such as video games, where AI algorithms have beaten every world champions, even in DOTA 2, considered one of the most complex video game in the industry.
The AI has achieved other good successes in fields like facial recognition, vocal recognition, image recognition, video recognition, translations or autonomous driving systems but with some percentages of error.
In open environments like social media or big data, AI’s modern algorithms have performed less, or sometimes AI’s results are dangerously wrong.
A scary example on how manipulable data, fake news and trolls could manipulate AI results was in 2016 on Twitter, Microsoft created a bot user, driven by AI algorithms, and after 24 hours start twitting that cultural minorities were to be killed and some other unpleasant statements.
Artificial intelligence makes sense, in environments with trusted data, and this concept is more important than the algorithms. AI-based system can really be helped by blockchain technology, to be trained in clean environments with validated and not manipulable data.
The concept of decentralized Artificial Intelligence is designed to solve the validation of AI’s algorithms, a crucial point for decentralized applications and at the same time to validate both the data it takes and data it generates.
The decentralization of the artificial intelligence algorithms solves a further problem, the computational power necessary to process the algorithms and perform complex calculations. The computational power needed to run Machine Learning algorithms is expensive and available only for large companies. In a distributed computational power network, the costs will be reduced exponentially allowing every company to use algorithms for complex calculations.
One of the most ambitious projects to decentralize artificial intelligence is SingularityNet, which uses Blockchain technology to decentralize artificial intelligence algorithms and at the same time a marketplace of AI to incentive developers and Data exchange.
Why could AI be the heart of the validation process for Blockchain?
AI in Blockchain technology can solve the dramatic problem of validation techniques.
The Proof of Work creates economic disincentives to fraud the network for validators (Miners) when they are solving complicated problems of decryption in exchange for cryptocurrencies. This Trustless methodology is based on the costs of the consumption of GPUs and electricity, making it unsustainable in terms of environmental impact and scalability, because it becomes more expensive for every validated block.
The current most trusted solution to solve the limits of the Proof of Work, promoted by Ethereum, is Casper. The concept is that a validator must block some cryptocurrencies (the stake), the Stake becomes its voting power, and if it tries to defraud the network, the validator loses the entire Stake. Casper discourage validators to fraud the network by burning their Crypto-Tokens rather than with high electricity/GPU’s costs.
Casper is a consensus method in the category of Proof of Stake algorithms. Another interesting POS method is the Delegated Proof of Stake (DPOS) developed by Daniel Larimer and the core of Bitshares, Steam it, EOS, and some other projects.
The DPOS validators are not disincentivized by burning their Tokens, but in contrary by incentives to be honest in the long-run. The POS allow Blockchain networks to scale and be more sustainable, but in the long-run, this consensus method is dangerous, because can generate a lobby of high stacked validators, with an enormous decision potential centralizing the network.
One of the most impressive consensus algorithms, promoted by AION and it’s called “Proof of Intelligence” which is similar to the Proof of Work but based on artificial intelligence.
This concept allows validators to use their computational power by training an AI to solve complex mathematical problems; this shared AI can solve cryptographic problems faster than a single miner in the POW. In the POI the shared AI continues to improve its solving skills, so when cryptographic difficulties increase, validators don’t need to spend more in therm of their computational power.
The Trustless data value
The real opportunity behind this technological challenge is the inestimable value that decentralized data and decentralized communities will have in the upcoming decade. The future of the merger of AI and blockchain technologies will be an interconnected internet of validated data, where both providers and users will be encouraged to use the most decentralized and validated platforms possible, because the value of their data will be calculated through how the data are tracked and validated with a continuous exchange, something like a “Validated Data Economy”.
For this purpose there is already a race to find the best method to calculate that, through projects like Nebulas, that reward dapps developer for the quality of their code and users engagement. Who’ll win this race will become the google and the exchange of decentralized data.
The motivation that is pushing this technology to a supersonic speed in terms of research, development, innovation and ideas, is an exponential economic advantage given by the use of Utility Tokens, crypto assets dedicated to projects, this new non-financial value is solving one of the biggest problem that the tech industry has today.
In the Technology Industry, the capitalization of a company is driven by intangible assets such as communities, algorithms, partnerships, and data. All of the methodologies that we use today to calculate these assets are rudimentary.
The Utility Tokens, without any connections to financial assets or speculative instruments, if it used as the key to interacting with the technology and data of tech products, can become a validated valuation tool for these intangible assets, based on their capitalization driven only by supply and demand in the market.
For example, if we have to calculate the value of data of a social network like Facebook, we wouldn’t have a scientific and validated model.
If Facebook would create some interaction’s tools with its database and the only way to access to this informations would be the Facebook cryptocurrency, the Market Capitalization of that Token, driven by supply and demand, could be a trusted and validated valuation of this social network’s data.
This new tool is a revolution for the tech industry, as well as the reason why the race to validated data will be the fundamental element of the multi-billion dollar companies of this technological wave.
From Vitalik Buterin:
“This, basically, is tokens 1.0. There are some good ideas, there are a lot of very bad ideas, and there’s a lot of very, very bad ideas, and quite a few scams as well. I expect that tokens 2.0 and the kinds of things that people will start building in 2018 and 2019 will generally be of substantially higher quality. Especially once we start seeing what the consequences of the first wave of tokens are in the medium-to-longer term … What is a good role for them and what is a role that doesn’t really make sense?”