The Holy Trinity of Disruptive Technology: Blockchain, Internet of Things and Artificial Intelligence. A critical perspective on how they shape our future.
“A year spent in artificial intelligence is enough to make one believe in God.”
Proclaimed Alan Perlis, a prominent 20th-century American computer scientist in his book ‘Epigrams on Programming’, believing that Artificial Intelligence (AI) will never surpass humans in terms of sophistication and thought process. Little did he expect that 36 years on, AI, together with blockchain and the Internet of things (IoT), has now formed the holy trinity of disruptive technology. Like the robot in the painting March of Intellect by Heath, they are now challenging our understanding of our world and giving rise to profound ethical questions on how they will impact on our future. Indeed, from Talos (the idea of an intelligent robot from the Greek Antiquity) to Google Duplex, an AI assistant which could provide a nearly flawless imitation of human sounding speech in 2018, technology has improved so rapidly that it is changing almost every aspect of the world as we know it. This article will serve as a guide to facilitate discussion of the combined effects of the the, barriers to their adoption and their latest developments.
Synergies within an Organism: Brain, Central Nervous System and Memory Network
If we consider the trio of technologies as parts of an organism, IoT is the human nervous system with a central function of feeling and collecting information. With improvements made in the field of sensor technology, it has the potential to transform the world of things into a world of data. AI would be the brain, capable of processing information collected through machine learning in a fast and effective way. Last but not least, with its ability to store and process information securely, blockchain would form the memory. It can also be explained using the concept of a ‘hive mind’ or collective consciousness - a social theory by Durkheim which suggests the ‘totality of beliefs and sentiments, common to the average members of society, forms a determinate system with a life on its own’. Indeed, with the vast amount of data we collect today, blockchain is able to integrate diverse data sets and provide a platform which draws consciousness to help with decision making and the functioning of a group. The wide-scale adoption of these technologies could only happen if their key traits and weaknesses are considered in tandem with each other. In the example of autonomous vehicles, blockchain could introduce security and transparency to the vast amount of data collected and analysed by IoT devices and AI can transform them into useful insights for users and businesses. Adopting these technologies as interoperable parts of an organism can help adopters achieve the data flywheel effect - when algorithms continue to improve with more data collected - at a much faster rate, accelerating the process of scaling.
Disrupting Healthcare- AI Revolution and Internet of Medical Things
In the healthcare industry, a large amount of data is currently processed and regulatory requirements, authenticity trust issues, and security concerns are some of the major barriers to information sharing. Integrating the trio of technology can ensure better collaboration and more accurate diagnoses and effective treatments to deliver cost-effective care.
Blockchain technology can remove the problem of unclear data ownership, improve data integrity and peer to peer accountability. Given the immutability of data, individual identities will remain secure while health data is added to the blockchain. On a blockchain platform, data can be stored in the existing databases or on cloud computing systems. A source can be created for each strand of data and redirected to the blockchain along with the patient’s public ID. Through an API, healthcare stakeholders can query the blockchain that provides the location where the data can be found without revealing identity. Subsequently, smart contracts are used to manage patient data access.
As a secure and easily authenticated platform, blockchain also allows for the integration of data collected from the Internet of Medical Things (IoMT) such as wearables like fitness trackers — this allows patients to generate authenticated and secure health data from multiple devices to monitor and access their daily routine. Once the data is collected by IoMT, AI system can rely on its self-correcting abilities to increase the accuracy of its feedback, reducing diagnostic and therapeutic errors and to tailor health and exercise plans, which can be adjusted as data are interpreted using machine learning algorithms.
Theresa May recently announced an AI revolution which aims to use AI to diagnose prostate, ovarian, lung and bowel cancer to prevent cancer-related death by examining medical records, habits and genetic information pooled form health charities. Artificial intelligence is also able to analyse healthcare data in a different format. An MIT-research already developed a machine learning algorithm which can analyse 3D scans up to 1,000 times faster than what is currently possible. This near real-time assessment can provide critical input during operation.
The combined use case of the trio in the healthcare industry is a testament of the clear benefits of integrating these technologies in terms of improved cost performance of computing, storage and bandwidth which enables the creation, sharing, analysis and secure storage of information without minimal human intervention.
Potential Barriers to Adoption
Despite the clear promise of value, successful adoption in the healthcare industry, and a large amount of investment made in the field, structural and behavioral barriers still prevent the trio of technologies from being adopted at scale.
Last month, Gokce considered Moore’s adoption curve and the Gartner Hype Cycle in mapping out blockchain dissemination. This time I broaden our perspective to include AI and IoT, and use the Sigmoid function as the benchmark, to offer you a different perspective on the structural barriers to the adoption of emerging technologies.
The lifecycle of technology, which is illustrated using the Sigmoid function (or the S curve) in Exhibit A, which would then inform our understanding of the potential barriers of adoption and the differentiated geographical impacts. According to MIT professor Erik Brynjolfsson, the s-curve patterns are usual for transformative, general purpose technologies like the steam engine, where the deployment lags are usually longer given that full benefits can only be attained with a high number of complementary co-inventions and investments, as well as business and regulatory models. Take the internet, for example, the inception to widespread usage took almost 2 decades to achieve. The number of users also followed an S curve pattern and follow the 4 stages from pioneering to mature.
As for blockchain, IoT and AI, they are currently somewhere in between pioneering and growth stage, although some application such as AI in self-driving care has already passed the pioneering stage, into the later stage of growth. However, widespread adoption would still require further experimentation, as well as development of new business strategies and policies, which could further prolong the adoption process.
In terms of behavioural barriers to adoption, the overall failure to realise value can be traced across many types of organisations in various stages of technological maturity in which they are only passively exploring without a clear direction and commitment. On the micro scale (Exhibit B), a common barrier to technological transformation is myopia in which businesses simply lack the foresight to fully comprehend the benefits of technological transformation and are unsure of the economic rewards relative to the risk of implementing new technologies at the initial stage of development.
However, even as the number of use case increases, a new barrier arises, which could be best explained by the concept of the ‘innovator’s dilemma’, in which investing in transformative technologies could disrupt existing business model, resulting in product/service cannibalisation. This inertia can further be felt as the benefit of the new model improves exponentially, while the existing business model struggles to stay relevant. In the final stage, when the distributive technology has brought about disruptive impacts, it reaches a point where the incumbent has to transform or cease to exist.
When considering inter-firm relationship from the broader industry’s perspective, the inability for the trio of technology to be adopted at scale can be explained by the cooperation/competition paradox in which companies are unwilling to cooperate in competition for market share.
Take cloud technology for example, when the two tech giant IBM and Dell finally decided to cooperate after a long stalemate, the industry achieved a breakthrough in terms of the technology progress despite the fact that that market share of each of the companies decreased.
As a result, instead of having a bigger share of a smaller pie, they now each own a smaller share of a much larger pie. Given the nature of the trio of technologies which derive value from large scale adoption, the industry is increasingly seeing a need to establish consortia - a multi-party group - to deliver solutions that work for multiple participants.
Shaping the future — Comprehensive Roadmaps and Hybrid Ecosystems
To fully harness the capabilities of the trio, it is important for businesses to not just blindly pursue the trend, but evaluate the key traits of each technology to determine whether it is really the solution to a given problem. It is also important for businesses to recognise the disruptive potential and look to the horizon to find strategic opportunities. Each technology has to be evaluated as far more than a trend on the hype-cycle (Exhibit C) but based on the potential to evolve not only as foundational components of enterprise IT but of corporate strategy; businesses need to develop a comprehensive road map and long-term goals related to industry trends, to build a hybrid ecosystem which allows it to create new capabilities and taking advantage of an active reshaping the existing one.
Incumbents need to reinforce its strategic ambidexterity and actively explore the new and exploit the old at the same time.
To build a functioning hybrid ecosystem, one which combines digital infrastructure and physical assets, incumbents need to reinforce its strategic ambidexterity and actively explore the new and exploit the old at the same time. To do so, a balance between creating new capabilities and taking advantage of and actively reshaping new ones.
Successful example includes GE and Simens, which launched Predix and Mindsphere- cloud-based platforms which utilise IoTs to improve the performance of industrial machines by analysing the data collected- respectively. When deployed strategically, the value of technological integration is almost immediate — Mindsphere helped Amtrack reduce delays by 33% only a year after its launch.
Moreover, hybrid ecosystems require dynamic business models capable of encompassing multiple stakeholders. As such, developing a replicable framework and an emphasis on credibility and collaboration is especially crucial. For example, Source AI, an advanced analytics tool by BCG GAMMA, is developed with the main aim of creating an enterprise-grade solution that would enable AI to be implemented at scale. As newcomers are increasingly placing themselves at the intersection of both the physical and digital, pressure for traditional players to transform digitally has never been greater.
To develop a comprehensive technological transformation agenda, companies should adopt framework which finds value — such as financial metrics and operating metrics to evaluate the extent of cost reduction and risk management deployment to explore revenue and innovation potential. Furthermore, companies should also develop their shaping capacity by investing in management skills essential for effective ecosystem building. In terms of consumers, while integrating technology into their strategy, businesses should aim to maximise transparency, enhance user trust and develop security that is both preventative and responsive based on customer lifecycle. It is important to recognise that given the disruptive potential and the trio of technologies, every sector will eventually become digital. To stay relevant, businesses need to take the leap of faith - even if it means cannibalising existing business models.
Technological Singularity — Impact of AI on Ethics and Legal Framework
The rapid deployment of these technologies also introduces new ethical concerns especially concerning bias in the field of artificial general intelligence and the algorithmic bias in machine learning models. Kurzweil, Hawking, and Musk are amongst many who have expressed concern over ‘the singularity’ - a point where the machine would outsmart human beings - and the potentially complex implications it has on our world. The hypothesis of the creation of ‘artificial superintelligence’ has led to calls for a more ethical implementation of AI systems and to imbue autonomous systems with a sense of ethics. However, this has proven to be a difficult task in its implementation and led raises questions as to ways in which fairness can be taught and racial and gender biases overcome.
Moving forward, AI researches and ethicists need to formulate ethical values as quantifiable parameters which can provide machines with explicit answers to potential ethical dilemmas. While it is a challenging task, a potential approach is to crowdsource data on explicit ethical measures to train algorithms. MIT’s Moral Machine is a project is an example of ways in which crowdsourced data could be used to help machines make better moral decisions in the context of self-driving cars.
While it is important to define ethical parameters for machines, policymakers and industry experts should also make AI systems more transparent. OpenCog and Open AI are examples of open source database for AI development. However, making a code open source does not make it comprehensible as neural networks are simply too complex to be scrutinised by human inspectors. Instead, the focus should be the process in which engineers quantified ethical values before programming them, as well as the outcomes that the AI has produced as a result of these choices.
What does the future hold? Synthetic Intelligence and Ambient Computing
Moving forward, the adoption and advance of the trio would only increase exponentially. In terms of AI, besides image recognition and data processing, progress in synthetic intelligence means that it could soon generate knowledge of this own. Beyond policy and regulation, this raises significant philosophical and epistemological questions about agency and the increasingly dialectical nature of our interaction with machines and computers. As for blockchain, use cases have become more diversified, ranging from healthcare to the arts, supply chain management to retail. Progress made in ambient computing is also elevating IoT beyond the mundane task of collecting information, shifting the focus to business process and model transformation — cities and industries are becoming increasingly connected. As a collective, the trio is disrupting existing economic frameworks and fundamentally transforming our relationships with machines. These exciting new developments call for all stakeholders - not just tech leaders, ethicists and regulators but also you, the reader - to come together, to envision and build a digital hybrid ecosystem in which data can be generated, collected, analysed and turned into insights, to shape our collective future.
London Blockchain Labs (LBL) is a non-profit made up of students, academics, businesses & policymakers who are passionate about blockchain technology and its adoption, based at UCL, LSE, Imperial College and London Business School.
LBL is hosting Europe’s largest university-led blockchain conference — Blockchain in the Digital Economy (BIDE) at Imperial College London covering the key trends across academia, government, and industry for 2019. Hear from world-leading experts and discuss blockchain and tech with UK’s tech talents from leading universities. Learn more and secure your place!
Author’s note: many thanks to Ilya for his guidance and feedback as I wrote this article. If you enjoyed this piece and would like to find out more about blockchain adoption, please follow our Facebook page.
If you’re interested in AI, technological transformation and the broader ontological debate about our interactions with the physical world, follow me.
