Exponential technologies like Artificial Intelligence, Blockchain and the Internet of Things (IoT) are creating significant breakthrough today. Lots of hype are woven around these technologies, huge investments are flowing into them, and new use cases are getting created on a regular basis. However, most organizations tend to study and assess the impact of these technologies in isolation, rather than as a combined force multiplier. This paper discusses how AI, Blockchain and IoT can address each other’s limitations, and enhance the overall value creation possibilities.
[siteorigin_widget class=”thinkup_builder_divider”][/siteorigin_widget]
Combinatorial Evolution
Combinatorial Evolution is the phenomenon of radical innovation and critical inventions created by the intentional combination of multiple technologies. It was a concept developed by W. Brian Arthur, an economist specializing in complexity theory. My view of combinatorial evolution is that it is a prime example of systems emergence, where linear and non-linear components interact, at regulated levels, to produce output with high non-linearities.
Artificial Intelligence, Blockchain and IoT are at different stages of maturity today. Artificial Intelligence, as a composite set of multiple technologies, has been around for several decades. However, it was only in the late 2000s that AI witnessed a tremendous leap after the advent of robust Deep Learning frameworks. Blockchain and IoT are relatively much younger. Blockchain was invented in 2008, and has gained a lot of traction since 2014. While the concept of connected devices is not new, IoT really came into force as an emerging technological force after the 2008 global recession. These three technologies, along with Big Data and Cloud, are the strongest pillars of the Fourth Industrial Revolution. Their convergence has already started showing signs of massive innovation at a scale and pace never witnessed before.
Enhancing Blockchain
Blockchain or Distributed Ledger Technology (DLT) is plagued by several problems today that often make large-scale production implementation infeasible and/or inefficient. Artificial Intelligence and IoT can play help in addressing some of these problems.
1. Security vulnerabilities like the 51% attack, fork issues, or code flaws in smart contracts can be easily exploited by hackers. AI can play a key role in addressing Blockchain security issues through Anomaly Detection techniques.
2. One of the key Blockchain production issues is that the ledger/database size grows over time as transactions get added, leading to performance problems and computational delays. This can be addressed through intelligent data pruning or smart ‘small ledger’ generation, both of which are powered by AI.
3. Blockchain applications are highly resource (cost, energy and infrastructure) intensive. Predictive and Prescriptive modeling driven through Machine Learning can enable optimal resource allocation and utilization.
4. Storing data in a ledger is not enough; generating insights is equally important. Data mining features can be developed in Blockchain through AI. Additionally, new capabilities like ‘cognitive’ consensus generation can be built, thus improving Blockchain’s overall efficiencies.
5. A distributed database system will always benefit from faster data accumulation and processing. This is how IoT’s Edge computing can enhance Blockchain, and make it more attractive for adopters.
Enhancing Artificial Intelligence
Artificial Intelligence is the new electricity, and almost every other company is in a race to extract value from it. However, we are still far from being able to optimally and efficiently create large-scale AI production solutions for solving critical business problems. Blockchain and IoT can help address some of the current limitations of AI technologies.
1. One of the biggest factors that stand in the way of greater AI adoption is the ‘Black Box’ nature of Deep Neural Network (DNN) models. It is very difficult, almost impossible, to accurately interpret why a certain prediction was made or why a certain action was taken by a DNN model. This has been a hard problem for AI researchers over the years. Some efforts have been made in this direction, e.g., the LIME (Locally Interpretable Model-Agnostic Explanations) framework. But it is still a major problem and that’s where Blockchain can play a key role. Cryptographic hash functions, vastly improved through recent advancements in Blockchain technology, can be created by models to explain most of their actions. These explanations can also be analyzed and leveraged to improve the internal functioning of AI models over time, thus enabling the creation of more efficient self-learning agents and systems.
2. Blockchain can also enable the fulfillment of regulatory compliance needs of AI systems and Bots. The identity of each AI agent can be effectively captured, and regulations can be enforced through Blockchain’s consensus feature.
3. Most AI solutions are centralized today, thus making them less effective in environments where decentralization and distributed computing may be needed. Decentralized-intelligence-at-scale, enabled through Blockchain and IoT’s edge computing, will offer optimal solutions for such cases.
Enhancing IoT
The Internet of Things is transforming the way devices and machines communicate with one another, and significantly adding to the benefits that get accrued from such communication. Advances in embedded systems & sensors, network infrastructure, Big data and Cloud technologies have provided a significant boost to IoT efforts in the past 2-3 years. Industrial IoT (IIoT) is the leading contributor to Industry 4.0. However, the adoption rate of IoT is still low due to several limitations.
1. The main concerns with IoT today stem from the security vulnerabilities of public networks that IoT devices leverage. Blockchain cryptography can enable trust among multiple devices, thus addressing some of these security concerns.
2. Artificial Intelligence can enable the development of ‘cognitive’ micro-processors and micro-controllers, thus paving the way for smarter edge computing through AI-enabled IoT.
3. IoT intermediaries and high settlement times are still big issues. Blockchain can help remove, or at least reduce the number of intermediaries, thus bringing down costs and settlement cycle times.
4. Many IoT solutions lack interoperability, and end up creating data silos. Blockchain’s ledger/database technology can enable the creation of unified data systems, which can then be effectively mined through AI.
[siteorigin_widget class=”thinkup_builder_divider”][/siteorigin_widget]
The State of Convergence today
Innovative organizations are gradually realizing that individual technologies are only discrete elements in the entire array of innovation, and path-breaking advancements can take place only when these discrete technologies converge in a systemic manner.
Some of the top companies have started building practical use cases around convergence. A good example is Fujitsu’s innovative solution to detect heat stress of workers through IoT sensors, take precautionary steps leveraging Machine Learning, and automatically pay insurance money to affected workers through Blockchain’s smart contracts.
There are four key factors today that act as barriers to migration towards full-scale technological convergence.
1. Convergence, as an area of study, is quite under-researched, which hampers development. Additionally, technical feasibilities are still not completely explored even in known use cases. Most ideas are generally in early proof-of-concept (POC) stage.
2. Scams in emerging technologies (e.g. Pincoin, Centratech, Plexcoin, Magos, etc.) often end up discouraging companies from investing efforts and funds in convergence projects.
3. Ethical questions like ‘Should we change or forget data on a Blockchain?’, ‘Should we use AI for certain industries like Defence’, and many others are still unanswered.
4. Artificial Intelligence, Blockchain and IoT are all evolving technologies. They need to go through their own maturity cycles before optimal benefits can be obtained from their convergence.
Closing Comments
The convergence of advanced technologies hold the key to massive-scale innovation in short cycle times. Each individual technology will go through its own maturity cycle. Over time, some technologies will become more robust, and some will fade away. While this paper has been about Artificial Intelligence, Blockchain and IoT, other emerging technologies (e.g. 3D Printing, Quantum Computing, etc.) will also be important. The first challenge for innovators is to mature these technologies. The second (and bigger) challenge is to integrate them in an efficient and scalable manner.
We are witnessing a slow but continuous increase in the number of convergence POCs. Established companies as well as start-ups are trying to solve different aspects of the convergence puzzle. Investments have started to pour in, academic labs are contributing to theoretical research in this area, and relevant talent is slowly being developed. All these are small steps but great signs.
The next 3 to 5 years are going to be really exciting!