Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Blockchain, AI, and IoT are all technology disruptions: Blockchain promotes decentralized applications in an open-data environment,  AI provides intelligence through a centralized data and analytics platform, and IoT enables devices/machines to be connected via a centralized cloud storage and processing service.

In the previous chapters, we discussed AI and its influence and the potentially explosive value generation if combined with IoT, RPA, and IT operations. In this chapter, we discuss another aspect of AI in a highly decentralized environment and how, if combined with blockchain technology and IoT, it can deliver far-reaching and amplified outcomes.

First, let’s have a quick overview of blockchain.

Blockchain Primer for Managers

We have witnessed two waves where initially scarce and expensive resources became cheap and technologies then emerged to exploit these resources. When computation (transistors) became cheap, PCs (personal computers) emerged; when bandwidth became cheap, the Internet emerged. Similarly, blockchain is the third wave where cheap storage allowed us to create a distributed (distributed ledgers to record ownership), open to all (yet secure) way of recording important information, thus introducing a robust, transparent, and trustworthy mechanism to issue and transfer assets in the virtual world.

Blockchain (distributed ledger) is a secure, distributed, and immutable database, accessed by all participants in a distributed network where transaction data is recorded (either on-chain for basic information or off-chain in the case of attachments) and audited. The data is stored in structures called blocks. The blocks are connected to each other in a chain through a hash (each block also includes a timestamp and a link to the previous block via its hash). The blocks have a header, which contains metadata information and the content (real transaction data). Since every block is connected to the previous one, as the number of participants and blocks grow, it becomes extremely difficult to modify information in the block chain without obtaining the network consensus.

Blockchain has three major characteristics:

  • Decentralized control : Blockchains operate on a shared control framework wherein it offers a common mechanism for participants for a common cause. No single entity in a blockchain network has exclusive rights granted to it that others do not have. This allows for the smooth formation of decentralized networks for various transactional services. This is highly influential in cases where best practices need to be made aware to everyone so as to improve the process as a whole, make it beneficial to the end user, and make it easy for all to ensure better governance and compliance.

  • Assets exchange : This is perhaps the most important feature of blockchain, since all transactions are recorded and no single entity has total control over what happens in the network. The system allows for transparent exchange of value items, which are data units that have value, like digital currencies, energy unit data, stock unit data, land deeds, educational certificates, and much more. It becomes easier to maintain in a trusted network critical information exchange and can revolutionize the way in which transactions are carried out with blockchain technology compared to traditional mechanisms. More reliability, security, and transparent control will ensure that transactions are viable under all circumstances and do not violate ethical policies.

  • Immutability : A transaction, once it’s recorded, is permanent and can’t be erased from the network. There is a permanent digital audit trail that lets everyone in the network see who did what. This forms the basis for smart contract management when contractual obligations become easier to validate from all parties involved. This will ideally mean complete elimination of fraudulent practices and keeping a check on all transactions carried out on such networks. It will become impossible for someone or a group of people to influence the entire network to commit a faulty transaction. No matter how long the network exists, all previous actions would be available to monitor and inspect for all members and stakeholders involved.

There are two methods applied to validate the transactions in the network.

Proof-of-work asks the participants (called miners) to solve complex mathematical problems in order to add a block. When you want to set a transaction, this is what happens behind the scenes:

  • Transactions are bundled together into what we call a block.

  • Miners verify that transactions within each block are legitimate. To do so, miners solve a mathematical puzzle known as proof-of-work problem.

  • A reward is given to the first miner who solves each blocks problem.

  • Verified transactions are stored in the public blockchain.

Proof-of-stake instead tries to attribute more mining power to participants who own more stakes (coins). It is still an algorithm, and the purpose is the same as of the proof of work, but the process to reach the goal is quite different. Under this system, forgers (the PoW equivalent of a miner) build blocks based on their stake in the blockchain’s network.

Blockchains follow two approaches to grant access to participants in the network:

  • Permission-less blockchains : Anyone can join the network and participate in the process of block verification to create consensus and also create smart contracts. Bitcoin and Ethereum blockchains are a few examples, where any user can join the network and start mining.

  • Permissioned blockchain : Only a restricted set of participants have the right to validate the block transactions. A permissioned blockchain may also restrict access to approved participants for creating smart contracts. Examples: R3 (Banks), EWF (Energy), and B3i (Insurance).

What happens when we merge IoT and the blockchain into a single, powerful platform?

Blockchain and the Internet of Things (IoT)

In principle, it makes a lot of sense, and there are several clear advantages of the idea of machines communicating with each other and humans and operating in a secure, trusted way via blockchain. IoT is all about continuously providing information about the state of the machines, Blockchain is all about an encrypted, distributed transaction recording system. Put them together and in theory, you get a verifiable, secure, and permanent method of recording data generated by smart machines.

What are the real-world benefits of blockchain and IoT convergence?

  • Solving the issue of oversight : In a connected world where there are multiple networks owned and administered by multiple organizations and vendors, it is important to have the oversight of what is happening across the entire network and when. In this type of distributed operations scenario, a permanent immutable record means custodianship and traceability. In the real world, physical goods passing between points in the supply chain is a classic example of loss of oversight. IoT devices continuously transmit data and blockchains register every single change to the state of goods in transit, which are visible to anyone authorized to connect to the network. This solves the problem of lack of oversight. If something goes wrong—breakages occur, theft, pilferage, etc.—then the blockchain record would make it easy to identify the weak link and raise alerts for action.

  • Establishing trust: The use of encryption and distributed immutable storage means that the information recorded in the blockchains at any point in time can be trusted by all parties involved, including machine-to-machine interactions. The “smart contract” facility enabled by blockchain networks helps establish agreements that will be executed when conditions are met. This is highly useful when it comes to, for example, authorizing machine-to-machine interactions and task executions when conditions indicate that tasks in the previous step are completed. IoT-enabled devices will securely interact and exchange information without any human involvement.

  • Ensuring heightened security : Both in the industrial world and consumer world, much of the data generated by IoT is highly sensitive and mission critical. This data needs to be shared with other machines and services in order to be useful. But it also means that this data is vulnerable and open for hackers to potentially misuse. This is where blockchains bring in robust access control, encryption algorithms, and security mechanisms that are difficult to breach.

What happens when we merge AI and the blockchain into a single, powerful offering?

Blockchain and AI

There is a bit of AI in everything around us, be it searching for something on Google to buying something through Amazon, commuting to work, playing our favorite playlist, monitoring health conditions, ordering groceries, etc. Our decisions are being influenced by algorithms that have been developed to cater to almost any kind of human persona and needs.

Blockchain, on the other hand, provides a way to exchange value-embedded data without friction and with trust, thereby making the data secure and ensuring that it is used for the intended and previously agreed upon purpose—nothing more, nothing less.

At a broad level, there are several ways in which AI and blockchain can complement each other.

AI Needs Encryption

Data on a blockchain is highly secure owing to the cryptography technology record transactions on blockchain. Blockchains are ideal for storing highly sensitive data through private keys in order for all of the data on the chain to be secure.

There is also a larger need for AI algorithms to work on the data, including the processes while it is still in an encrypted state, because any phase of the process in which the data is unencrypted represents a security risk.

Blockchain Can Help AI Become Explainable

AI algorithms analyze a large number of variables to “learn” patterns hidden within a vast amount of data and arrive at predictions. The learning process and the complexity of algorithms often becomes incomprehensible to end users—why the algorithm recommended what it recommended, how the algorithm arrived at the recommendation, which variables it used and which ones got dropped and why, and many more such questions.

AI offers huge advantages in many fields; however, if the outcomes are not explainable, then they won’t be trusted by the end user. Blockchain technology can help AI record the entire learning and decision-making process, thus providing a level of transparency and insight to the end users.

AI Can Make Blockchains More Efficient

Due to heavy focus on encryption, blockchains require large amounts of compute and processing power. For example, hashing algorithms to mine blocks on the blockchain effectively try every combination of characters until they find one that can verify a transaction. Sophisticated machine-learning algorithms can optimize the blockchain mining algorithms, if they are fed with the right training data.

AI Can Make Blockchain Scalable

Blockchain by design is immutable, which means managing growing volumes of data becomes a stiff challenge. “Blockchain pruning” (removing data about inactive or closed down transactions) could be a possible solution. AI algorithms can play a role in automatically archiving using data sharding techniques to make blockchain more efficient and scalable.

Integrated AI and Blockchain Value Propositions

Improved Business Data Models

While existing AI-driven business models rely on data produced by organizations and their ecosystem partners, there are drawbacks when it comes to transparent data sharing. Privacy concerns, abuse of data, and fraudulent contracts make it impossible to create fully open data systems. For AI to produce the most accurate decisions, it needs virtually no barriers to information access from the entire business ecosystem. This frictionless information access would be provided by blockchain technology, as all stakeholders would be more willing to share information since no single entity can take ownership of the network.

Data sources from all key players can easily be integrated with a common AI-powered analytics platform. The way the data is going to be used would be defined in the blockchain ledger for the whole transaction, so there is no chance of data owners being subjected to scenarios where they would lose the credibility of their data or their data would lose its value.

Once an uninterrupted supply of data from multiple stakeholders is established, AI systems can dive deep into them and study the patterns and other aspects to uncover insights that were never witnessed before. As the number of participants in the network grows, there would be more possibilities to extract more genuine insights and the systems can train themselves to better respond to scenarios based on how each stakeholder would respond, by identifying best practices and discovering the best solution to a new scenario.

Newer Insights to Discovery

With transparent information access, data silos could be more efficiently merged and this would help AI to maneuver through newer data combinations and discover new patterns. With blockchains it is easier to eliminate data manipulation, so AI can be used to create new authentic classifiers and filters for data, since it is possible to verify their authenticity on a decentralized blockchain framework.

This is particularly important in scenarios that do profiling and segmentations for better predictive customer engagement. More data sets would allow micro-segmentation and hyper-personalization.

Intelligent Predictions

AI has come a long way in helping us predict outcomes based on data collected from various sources. However, at times such predictions may become factually incorrect because of faulty data generation systems, fraudulent tampering of data sources, or loosely governed analytics methodology employed in AI systems. With blockchain it is easy to authenticate data sources and the analytical methods to be applied to the data, as they would have to agree to predetermined contracts and best practices. This paves the way for AI systems to work only on authentic and genuine data sets, thereby resulting in accurate predictions.

More genuine data sources would ensure that AI systems work more efficiently to mine insights. These insights would then be utilized by deep learning algorithms to arrive at more factually correct decisions. This would ensure better predictions of end consumer behavior for businesses or end user behavior for better governance processes.

Data Intellectual Property Rights

AI-enabled data models always create great success stories that inspire others. But very often creators of such models refuse to share their data model information for lack of protection and copyright policies. Even though there exist copyright policies, data models can be masked intelligently to showcase totally different architectures that can’t be identified even by the creators of the original data models. However, with the integration of blockchain technology, creators can easily share their data models without losing exclusive rights and patents on their discoveries. A tamper-resistant global registry can be maintained with your cryptographic digital signature that contains both your data and its associated models. Who does what and when with this data would always be available on the public domain, thereby making it virtually impossible for anyone to claim authorship of your creations.

With integrated AI tools, it is possible to analyze trends and behavior from these data models and enable powerful insights discovery from them. It also empowers businesses and content or data owners to claim their due privileges and rights to intellectual property for time immemorial. It offers an attractive way to monetize original data or content for professionals in multiple streams.

Autonomous Organizations

This is virtually the dream scenario of AI. An environment where machines perform tasks without human interventions. All that is needed is a globally verified working policy and set of instructions that need to be secure, tamperproof, and not owned by any single entity to set up a training and learning environment for AI-enabled systems. With blockchains, this is easier done than said and before you know it, AI systems can read data, process it, and operate your key business tasks. Chances of failure or faulty operational outcomes are negligible since the entire system is governed by smart contracts residing on decentralized blockchain networks. The network nodes store states, parameters, behaviors, results, etc., of the entire operations systems and allow AI systems to act accordingly to predetermined smart policies.

Such a level of autonomous operations has never been witnessed before; however, with AI-enabled data secure blockchain systems, this mode of autonomous operation is quite feasible. Deviations from predetermined behavior only occur when the system is asked to deviate by controlling nodes under the supervision of all stakeholders. Such a secure and transparent monitoring and control mechanism ensures that things do not go haywire at any point in time and human intervention is kept at a minimum, or in ideal conditions there will not be any involvement at all.

A few examples of AI and blockchain technologies coming together to deliver industry specific outcomes are discussed in the following sections.

Smarter Finance

AI and Blockchain technology are already having a significant impact on financial services on an individual basis. Their combination would result in even better transformative ways for financial entities to conduct transactional services. For example, forming banking consortiums to derive new service layers, policies to prevent fraudulent usage and access of customer information, etc. can all be achieved with AI and blockchain combined.

While the insights would be offered by AI, the underlying trusting platform would be blockchain. Banks would willingly contribute their data and information for a greater cause, as there is no single entity that has the ownership of the entire consortium. With such a secure and connected sharing ecosystem, AI systems can traverse a wide variety of demographic and market data, thereby uncovering new insights about business growth and potential opportunities to serve customers better.

Intelligent Retail

E-commerce and digital retail solutions for in-store sales have revolutionized consumer shopping experiences like never before. Today one of the biggest investments made by retail organizations on the technology infrastructure side are analytics and prediction systems. For a modern day retailer, getting to know a consumer’s buying decision before he or she actually makes purchases is the USP for successful experiences and checkout.

AI-enabled systems are already driving more sales for retailers with more intelligent prediction capabilities. However, there are still areas where AI is left to work on unverified data sets to arrive at decisions. Data about products or component quality from partners and or vendors is often not verifiable. If you provide the most suitable product but it ends up being a failure in terms of quality, then the entire retail experience is destroyed at that moment and could result in the customer leaving you forever.

With blockchain-integrated systems, it is possible to verify claims of vendors and suppliers regarding the quality and specifications of the product, as they would be available for inspection and analysis at any point in time. Accountability of all stakeholders increases and this collective responsibility results in only the best products reaching the shelf or on display in an online store. With more accurate data, AI systems can uncover new insights like, for example, new pricing points based on authenticity of materials (sourced from prime locations or exquisite places), and this leads to more value for creative marketing as well.

Hassle-Free Customer Loyalty Programs

Almost every business has a loyalty program (part of their customer relationship and marketing strategy), through which they reward their valued customers and encourage them to stay loyal (through differentiated treatment, disproportionate discounts, access to yet to be generally available products or services, coupons, vouchers, etc.). A good loyalty program is not only about rewarding valued customers, rather it’s an important strategy to strengthen customer relationships.

In reality, from a consumer perspective, we are subjected to multiple loyalty programs, each having its own rewards mechanism, maze of point systems, and cumbersome redemption processes. The result is that the consumer is not benefiting and the business is not enjoying customer loyalty in the true sense.

Blockchain may just be the answer. For consumers getting lost in the multitude of loyalty programs (physical cards or digital wallets), blockchain could provide a single unified platform that seamlessly manages each loyalty program option, including the limitations and redemption rules. How?

Blockchain enables a distributed ledger of transactions shared across a network of participants (consumers as well as businesses). When a new transaction occurs (a loyalty point is issued, redeemed, or exchanged), blockchain will issue an algorithm-generated token against that transaction. The tokens are grouped into blocks and distributed across the network, updating every ledger at once. In addition, blockchain ensures the blocks are validated and linked to older blocks, thus creating a secure and transparent audit trail of all transactions, without the need for intermediaries.

Customer loyalty programs on blockchain would mean a robust and trusted partner network for businesses, where they can collaborate and offer more options outside of their own core products and services, thus breaking out of the narrowly defined individual loyalty programs and reducing customer hassles.

Creative Passport

These are smart contracts for music industry. A major painpoint in the music industry is that the songwriters, producers, and musicians are always in the blind spot as far as how the copyrights are handled, how their royalty payments are calculated, and in general how digital products are managed and distributed and how the consumers are consuming their digital products.

This poses significant challenges to both the producers and the consumers.

Blockchain, with its capability to provide a verified, transparent, and distributed global registry of music for managing the digital product’s lifecycle, can eliminate the current creative industry challenges and at the same time provide a seamless experience for anyone involved in creating or interacting with music. For example, the activity of listening to a song would automatically trigger a chain of events in the blockchain system for everyone involved in the journey of that song with anyone who wants to experience the song or do business with it (an individual consumer, a digital service provider like Spotify or iTunes, a radio station playing the song, or a film production crew using the song, etc.). The blockchain system will have the added advantage of storing metadata about the music, sort of a “creative passport,” which could then be updated and accessible to everyone.

Transparent Governance

Every year, governments across the world spend billions on social improvement and citizen welfare schemes. However, due to corruption and inefficient middle layers and lack of proper auditing systems, a good portion of that money does not reach its intended recipients. Besides this, there is a huge waste of manpower involved in performing redundant and mundane tasks such as data entry, proofreading, verifications, log entries, etc.

By bringing in smart automated management systems powered by AI and built on blockchain technology, a high degree of efficiency and transparency can be achieved in governance mechanisms. Using AI, it is possible to free human resources for more important activities where all redundant tasks can be automated. A great deal of insight generation to see progress of the schemes and policy implementations can be created by AI systems that eliminate any form of human interference or tampering, thereby avoiding corruption.

With Blockchain, it is possible for citizens and governments to establish trustworthy policy implementation and data exchange mechanisms without worrying about faults and corrupt or malicious involvement. Citizens can verify claims of government spending and escalate concerns with proof. It is possible for governments to monitor all activities of all departments, thanks to the smart audit trails available in blockchain systems.

Globalized Verification Systems

Today there are a lot of verification systems across the globe that require information sharing and integrated view crossing boundaries, be it citizenship and identity verification, immigration details, credit and financial history, etc. This data is highly sensitive and can be manipulated if it lands in the wrong hands. However, for the smooth functioning of global systems, such as immigration, banking, educational, scientific research programs, etc., it is vital to have transparent information access for verification of claims and identity.

AI-enabled systems would help to study deep patterns stored in data about people and predict their credibility scores or financial health or identity, which is vital for security reasons such as in preventive measures for terrorism-related offences or financial crimes such as anti-money laundering, etc.

A globally decentralized registry for crime records, financial fraudulence, etc. would ensure that AI systems can train themselves of verified data to arrive at conclusions and spot outliers or anomalies. Blockchain-powered global verification systems would help to bridge the gap between global investigative and verification agencies, thereby smoothening verification processes across the globe and ultimately resulting in better cross-border commutes.

Innovative Audits and Compliance Systems

With blockchain creating a decentralized network for independent verification of claims, it is easier for authorities as well as businesses to adhere to compliance and regulatory requirements. Every transaction on a blockchain network leaves a digital trail that cannot be tampered with and is irreversible. This trail can be audited by any authority and can be used to verify claims of authenticity.

Using AI, data from these audits can be utilized to provide better services to customers and citizens, such as accurate pricing and legal information about real estate, tax calculations, and tax accountability, compliance to nationally or globally accepted quality standards for services and products, and much more. The possibilities are limitless as more authoritative data can help AI systems provide uninterrupted and factually correct information and services to end users.

Data and AI Monetization

Monetizing collected data (our data) is a huge revenue source for large companies, such as Facebook and Google.

With everything we do on the web or using apps, we leave behind huge data footprints. This data is aggregated and eventually used to cross-sell and up-sell stuff to us . This makes data incredibly valuable—more valuable than we the original data generators realize—and we are giving it away for free.

Blockchain enables microtransactions that make it possible for us to own, control, and monetize our own data. The same goes for AI. If we develop AI algorithms using our own data, we can put the algorithms on an AI marketplace for others to use. We are then paid usage based fees.

The Case for Decentralized AI

Until recently, the contemporary AI industry was based on a centralized development and usage model where machine learning solutions were designed, developed, and accessed as a part of cloud-based AI platform and associated APIs. Now we are moving toward the next frontier—decentralized AIs that can run and train on local devices or make decisions in decentralized networks like blockchain.

The existing AI market is increasingly controlled by tech giants like Google, IBM, and Microsoft, all of which offer cloud-based AI solutions and APIs. This model assumes little control of users over the AI products, and in the long run, such a centralized model could lead to the monopolization of the AI market. This could cause unfair pricing, a lack of transparency, interoperability, and limited participation of smaller companies in AI innovation. Fortunately, we are witnessing the emergence of a decentralized AI market, born at the intersection of blockchain, on-device AI, and edge computing/IoT.

The transition to decentralized AI is enabled by new technologies, such as Google’s federated learning, that allow for crowd-training of ML algorithms, device-centric AI that run and train ML models on mobile devices, and the use of AI in DAOs (decentralized autonomous organizations) on blockchain networks.

AI and Decentralized Autonomous Organizations

DAOs manifest when we entrust some or all decision-making responsibilities to AI agents. For example, If you are a holder of ownership rights, you can cede your decision making (e.g., yes/no votes) to an AI agent (another smart contract) that will make all the decisions on your behalf. Or, in a more radical scenario, we can entrust AI agents to play the role of managers and take all organizational and business decisions. For example, imagine an AI agent for marketing, where the AI agent selects the best ad companies to place your ads with. After each marketing cycle, the AI agent would assess the ROI and adjust the marketing strategy automatically.

In essence, DAOs take us to a qualitatively new economic reality where AI agents become business administrators and are overseeing business performance and continuously learning from and with other AI agents in a decentralized network. Terra0 is a project proposed by Paul Seidler and Paul Kolling from the University of Arts, Berlin. The concept involves augmented intelligence agents managing and monetizing a self-owned forest. Forestland ownership is structured as an DAO with smart contracts on the Ethereum blockchain. Then, using drones and satellites, the AI agents evaluate the woodstock and decide how much and when to sell in the market. Once the project is up and running, the DAO can pay out debts to its initial owners and eventually turn the forest into an autonomous, self-owned entity that controls its own resources.

Recent advances in decentralized AI have been made thanks to on-device optimization of AI/ML for smart phones and production of dedicated chips for mobile AI and for desktops (e.g., Google’s TPU).

Decentralized AI gained powerful momentum in April 2017 after Google announced its new Federated Learning concept. This innovation signals a transition to fully decentralized learning and device-centric AI where machine learning models are trained directly on user smart phones. Keeping the privacy of user data intact, Google can now outsource AI training to Android users, enabling on-device improvement of shared models. Federated Learning will solve the problem of high-latency and low-throughput connections where users have to connect to remote servers to use ML software.

The move toward device-centric AI can also be seen in the release of Google’s TensorFlow Lite, a mobile version of a machine learning library fined-tuned to the computational and power constraints of smart phones. In June 2017, Apple followed Google’s lead by releasing its Core ML library for iOS devices. The library ships with the optimized general-purpose ML models and tools to convert third-party models into the iOS format. Making models available locally without a network connection will make it easier to develop mobile applications with AI functionality.

In the long run, a combination of on-device AI and decentralized learning will make AI more democratic and widespread than ever before, resulting in more and more organizations becoming DAOs.

What is accelerating this trend?

  • Autonomous AI on blockchain: DAOs are algorithmic companies run by autonomous AI agents. When you put these autonomous AI agents on an Ethereum blockchain, you can effectively manage distribution of royalties, subscription payments, smart contracts, and more.

  • Decentralized learning at the edge: On the device, AI capabilities such as mobile machine learning libraries from Apple’s CoreML allows complex AI algorithms to run on IoT devices like sensors, security cameras, drones, or autonomous vehicles.

  • Decentralized intelligence networks : Decentralized networks built on the blockchain such as SingularityNET allow any company or researcher to monetize their AI solutions and get access to a variety of AI algorithms. SingularityNET enables cross-AI capabilities through protocols that support data exchange and sharing across different algorithms, which is helpful in building multitier and connected AI applications.

Such applications can combine multiple algorithms and perform different sub-tasks and then have access to the training data exchanged on the network. One example of such an approach is the development of a comprehensive AI-based cybersecurity solution. Currently, there is no single software package that handles all security-related tasks, which means that companies have to use various centralized AI solutions and customize them to their needs. Developing such a solution also brings about the problem of data security and domain-specific knowledge.

Homomorphic encryption that makes different data sets visible to different classes of users in different aspects can solve this problem. Companies can also combine the expertise of different cybersecurity AI agents on the network, which will safely exchange security information, outsource tasks, and cooperate in solving common security issues. Such decentralized network will involve efficient division of labor and offer access to solutions without having to manually obtain data and customize algorithms.

Decentralized intelligence has a number of advantages over centralized solutions in the following scenarios:

  • You need an autonomous AI solution that runs in the decentralized environment and implements contractual obligations: By definition, centralized proprietary solutions cannot be exposed to many users in the decentralized network. If your goal is to run a fully autonomous AI agent that makes smart managerial decisions and distributing profits, decentralized AI on blockchain is the way to go.

  • You need an AI optimized for the on-device performance and not dependent on network connectivity: Due to network connectivity problems, battery power constraints, and low computing power, mobile devices are not a good option for running cloud-based AI software. In particular, high latency and low throughput can compromise the speed and performance of AI applications, adversely affecting user experience. Whenever you need to run and train AI on mobile devices, a decentralized network becomes a better option.

  • You want to sell your AI algorithms while maintaining proprietary rights : There is no way to sell your AI algorithm and retain proprietary rights for it (the same goes for selling mobile apps in the app store right now). To fill this gap, decentralized AI networks offer an opportunity for developers to make their algorithms available for commercial usage as AI as a service.

At the same time, centralized AI still remains a good option if you need a very generalized ML model that you can easily plug into your application. Google, Microsoft, and IBM have developed the best generalized machine learning models on the market that are trained on huge data sets and built according to the top ML standards and bleeding-edge ML algorithms. Reinventing the wheel is not an option if you want proven image or speech recognition features in your application. A more viable solution is to use cloud-based ML APIs provided within a pay-as-you-go model, which ensures cost efficiency and scalability of your AI-based solutions. Major providers of centralized AI have a comprehensive suite of services for image and video recognition, emotion AI, speech recognition, predictive modeling, and other common AI/ML tasks.

In the long run, decentralized solutions can produce the radical democratization of the AI market, optimization of solutions for a wide variety of use cases, easy integration and communication between different algorithms through a single protocol, and the development of interoperability standards, which will ultimately lead us to the era of AGI (artificial general intelligence).

Conclusion

The combination of the blockchain technology and artificial intelligence is still a largely undiscovered area. Traditional AI methods follow a centralized development and deployment pattern. You have a cloud-based AI platform, you ingest data and do your predictive models and then, through APIs, access the models from various business processes or applications. There are a number of challenges to this approach: latency, data security, network bandwidth, performance, etc.

Now, imagine a highly decentralized approach where your predictive models run, train, and even make decisions on local devices in decentralized networks like the blockchain. This is decentralized AI!

There are significant benefits of decentralized AI over traditional AI:

  • Minimal latency (no dependency on a network connection)

  • Training is more efficient (done in a decentralized way)

  • Less power consumption (again, no dependency on a network connection)

AI can be incredibly disruptive and must be designed with utmost precautions. Blockchain can greatly assist with the cause. How the convergence between these two technologies will progress is anyone’s guess.

In this chapter we touched upon several disruptive technologies (blockchain, AI, IoT, and cloud), each extremely powerful on its own. When they are combined, they can deliver far-reaching transformative outcomes. In the next chapter, we touch upon the thorny topic of ethics and AI, which has raised enormous debates worldwide.

References

  1. 1.

    https://www.forbes.com/sites/forbestechcouncil/2017/11/16/why-decentralized-artificial-intelligence-will-reinvent-the-industry-as-we-know-it/#3bcedba7511a

  2. 2.

    https://blog.bigchaindb.com/blockchains-for-artificial-intelligence-ec63b0284984

  3. 3.

    https://www.forbes.com/sites/forbestechcouncil/2017/11/16/why-decentralized-artificial-intelligence-will-reinvent-the-industry-as-we-know-it/#3bcedba7511a

  4. 4.

    https://blog.bigchaindb.com/blockchains-for-artificial-intelligence-ec63b0284984

  5. 5.

    https://blog.bigchaindb.com/blockchains-for-artificial-intelligence-ec63b0284984