Blockchain Technology: the path to utopia or dystopia?
Ananth Natarajan — Founder, cybereum
Imagine a world where every activity and event can be logged on a tamper-resistant ledger irrevocably attached to an individual throughout their life. Where permanent, immutable digital and surveillance records are used to control a person’s position in society; where everything that one has said and done, with little context for those actions, will affect or even destroy a person’s relationships and professional or academic career; where society is governed by AI algorithms residing in distributed ledgers tapped into omnipresent sensors which log data to a permanent, indestructible ledger.
“Fate, it seems, is not without a sense of irony” — Morpheus (probably in the year 2199)
Ironically, this is a reality that we have made possible with a technology that was designed to do the exact opposite. Blockchain technology, like many tech developments, was born out of a utopian ideal. It came from a vision for a cashless society that maximized individual autonomy with freedom from oversight and taxation. However, the potential for blockchain to fetter us to a system using our records is innate to a tamper-resistant, extremely durable, and verifiable record. The case for blockchain is made in a myriad of articles. Here, despite being a blockchain company, we adopt a contrarian viewpoint and discuss the drawbacks of blockchain technology that come from its strengths rather than its weaknesses. The concern with the scenarios we outline is not that it will come true but that it can.
Blockchain technology enables secure, scalable ledgers with the properties of tamper-resistance, decentralized verification, and pseudonymity. A Blockchain is a distributed digital ledger of transactions stored in immutable tamper-resistant blocks. Each block’s fingerprint is present in succeeding blocks, chaining them together. A hash function that generates these fingerprints, or hashes, is also used to create puzzles used for mining in some blockchains and imparts tamper-resistance to the ledger. A hash can be easily verified to originate from specific data, but finding the same hash for different data is computationally infeasible. Users have two keys a secret (or private key) and a public key. The secret key is the key to ownership and is used for signing. A user’s public key is their identity on the blockchain. Anyone can verify user signatures using the public key without needing to know the secret key, enabling pseudonymity and conferring non-repudiation since a user cannot deny their signature. It is practically impossible to derive a private key from a public key. Thus, verifiability is imparted without requiring centralization, and pseudonymity is imparted by associating verifiability with public keys.
The potential for the internet and social media as powerful tools for control and pernicious purposes has been recognized and exploited by a myriad of state and non-state actors. Similarly, the potential for blockchain technology to control and limit individual freedom is starting to be recognized and will be exploited. Projects that begin with utopian ideals for decentralization and democratizing information often end up being dominated by groups of people with similar beliefs and high barriers for entry to outsiders in nominally egalitarian communities. This is illustrated by the current state of Wikipedia, where the final version of truth seems to be increasingly controlled by closed communities with a high bar for newcomers. Social media initially envisaged as a tool to connect people and communities across borders, has been used to divide people, effectively disseminate conspiracy theories, influence political outcomes, and recruit people for extremism (Al Qaida & ISIS made highly effective use of social media). Personal profiles scrapped from social media accounts, increasingly aided by artificial intelligence (AI), will result in far more detailed personal and even personality profiles¹ than the Stasi, the notorious East German Secret Police², ever dreamed of. The marriage of social media with blockchain will add the property of non-reputability and permanence to people’s digital traces. So, a good question to ask is if 2084 could look like a blockchain-powered version of George Orwell’s 1984.
A chain to find us all, bring us all, and bind us all
AI, with networked sensors and big data will be able to pierce through cloud, shadow, earth, and flesh, see all…and record it all in a tamper-resistant ledger. — with appologies to Sauron
Blockchain might well end up being the chain that binds us to the system. A future where personal records are permanently recorded in tamper-resistant ledgers has become possible with blockchain technology. These records, interpreted by opaque AI algorithms, can enable effective control, restrict access to opportunities, and affect the fundamental quality of human life. It only requires a few switches to be flipped to change blockchain from a libertarian to a totalitarian instrument.
Cryptocurrency mining theoretically democratizes transaction verification. But the vulnerability of cryptocurrency mining operations to centralization through majority control of the hash rate has been well known from the earliest days. The two main mechanisms for selecting validators to confer rights and rewards to create new blocks are Proof of Work (PoW) and Proof of Stake (PoS). PoW incentivizes the ability to commandeer large pools of capital for the farms of specialized hardware and huge amounts of energy required to solve hashing puzzles. PoS incentivizes the ability to control large amounts of tokens to stake. The probability of mining the next block in both mechanisms is related to the resources one can commandeer. The rewards that accrue from block creation add to the resources for commandeering hardware, energy, or tokens at one’s disposal, forming a vicious centralization cycle.
Furthermore, cryptocurrency mining operations have a large footprint both literally and in terms of energy consumption, as exemplified by the bitcoin mining industry; This makes them highly visible and vulnerable to government action, as was demonstrated recently (at the time of writing). While possibly required for investor protection and compliance, growing regulatory oversight poses fundamental questions to the privacy associated with cryptocurrencies and fosters centralization. Furthermore, we will see that pseudonymity is far from being a panacea for privacy. Blockchain technology and cryptocurrency may well be the best tools for profiling and control that governments and corporations ever had.
Many initiatives aim to put tracking systems, public records, academic credentials, and even social networks on blockchains to solve important real problems, but can create the links to bind us to the system. Supply chain blockchain startups are creating important tools for transparency, ethical sourcing, and traceability, but the same verifiable traceability can be easily extended to recording the movement of any goods, services, and also of people in tamper-resistant ledgers with granular and non-repudiable data. In blockchains such as bitcoin, the key characteristic that enables privacy is pseudonymity from identities being associated with public keys rather than individuals. However, when services and careers are tied to records on a blockchain, pseudonymity becomes redundant. Many systems may simply eschew pseudonymity, as with social credit systems using distributed ledgers. Networked sensors on blockchains, numbering in the billions and made much more powerful with 5G technology, along with AI applications such as facial recognition to feed tamper-resistant logs enabled by blockchain technology, are already being deployed. This is exemplified by a national social credit system currently being deployed. There is intense interest in blockchain from governments, not to promote individual autonomy or privacy, but from its potential for control.
Pseudonymity is the lynchpin of privacy; identification of real-world people or organizations will enable their non-repudiatable association with records on blockchains. But pseudonymity is far from a guarantee for privacy, as the ability to identify a person or entity behind a public key is inherent in the pattern of activity generated by an account. This was exemplified by the US Justice Department announcement of tracing most Bitcoins paid to hackers quickly after a famous 2021 ransomware attack. Indeed, non-repudiablity, provenance, and permanence made the transactions more, not less, traceable. Associating activity patterns with real-world identities will be made even easier with AI, which is good at pattern recognition. Also, while cryptography may be challenging to crack, something that advances in quantum computing will challenge, private keys are only as secure as the means of storing them. Traditional hacking techniques, and those with enough resources, can find ways of obtaining private keys to seize identities or funds.
Blockchains will become indispensable as they become central to the competition between companies and nations, especially in combination with AI, with AI increasingly edging out humans from decision-making loops. There is an increasing and highly sought-after convergence of blockchain and AI. Along with feeding off data and connected devices on blockchains, AI applications will increasingly reside on the blockchain. The undoubted utility of verifiable records will incentivize blockchain adoption, and network effects will accelerate this. Once in place, it may well be impossible to escape the clutches of a tamper-resistant and durable system enmeshed into society and the economy. We will end up in a situation where, to quote Kafka, “It is often safer to be in chains than to be free.” Blockchain may well make this more possible than any technology that has come before. We may have created and chained ourselves to a monster that will control us. As Dickens wrote, “We forge the chains we wear in life.”
Blockchain with AI — Limitations of cyberocracy
As we saw, blockchain is a dual-use technology that can enable its early vision of decentralization and autonomy but can also enable centralization and control to a degree hitherto unachievable. Centralization and control are not only a threat to individual rights and liberty. There are several issues with the use of blockchain and AI for social and economic control. Blockchain is great for records, but much real meaning lies in the context of the record. This context often lies in the time, place, and other determining factors. Recorded thoughts on social media or educational attainments do not capture individual constraints, evolution, and potential. Such context is in the realm of artificial general intelligence (AGI), which is extremely challenging and still fallible; some experts consider AGI unattainable³. Information alone is insufficient for objective decisions. It is now well known that human decision-making and data interpretation suffer from bias⁴. AI models, which tune themselves as they train from data, are known to suffer from bias as with humans, and attempts to de-bias AI can lead to different biases in unforeseeable ways. In addition to bias, cognitive limitations arise from the subjective nature of experience⁵. AI models, such as neural networks inspired by the human brain, suffer from many of the same issues, biases, and opaqueness regarding the inner workings. The effectiveness of AI neural networks is increasingly making them indispensable in an exponentially growing number of applications, but the models are effectively black boxes. They are unexplainable in several ways and may remain incomprehensible to deterministic understanding. This is made more prevalent by the commercial incentive to safeguard AI models.
Centralized solutions such as social credit, whose adoption can be turbocharged by blockchain and AI, are often ineffective and destructive in complex systems. Complex systems are characterized by emergent outcomes and chaotic behavior with inherent challenges to predictability⁶. Even if the interacting components of a complex system are known, it is practically impossible to determine all emergent outcomes from interactions⁷. The real world is characterized by a high degree of unpredictability and, conversely, by high confidence in prediction, especially by experts⁸. This may be said of both the current champions of blockchain technology and those who may try to use that technology to impose centralized solutions to complex problems.
The futility of centralized solutions when complex social and economic systems are involved was seen in human tragedies on a massive scale in the last century. These include millions of deaths resulting from forcible farm collectivization in the former Soviet Union and the attempts at top-down social and economic change in the People’s Republic of China. Planners could not foresee the consequences of centrally planned policies on complex systems and lacked the flexibility to recognize and respond to them. Problem-definition and goal-formulation are very difficult for social systems and often intractable due to their subjective nature⁹. Such “wicked problems” cannot have optimal solutions and are impossible to solve in a realistic time and resource framework, in contrast to “tame” problems in, for instance, chess. Eminent organizational theorist Russell Ackoff famously characterized complex interconnected problems as a” mess” that is more feasible to manage than solve¹⁰. This is reflected in the repeated failure of sophisticated economic models for real forecasting⁸. A discussion of the dangers and potential for governance enabled by control over information and networks, or cyberocracy, can be found in Ronfeldt, D. and Varda¹¹. The related threat from, and attraction to, Algocracy (literally, rule by algorithms) from AI is discussed at length by Danaher¹², whose paper is reviewed here by Vachnadze. Therefore, as seductive as the idea of combining blockchain with AI to plan and control social and economic systems is, it may be flawed and pernicious¹². Complex problems simply do not lend themselves to deterministic top-down solutions. Adopting this course with the power of blockchain records empowered by AI can lead to severe consequences over time. It can lead to a future that is far less free and eventually far less prosperous.
Dependence on AI applications distributed in blockchains to manage our civilization is dangerous; Cyberocracy and Algocracy can be far worse than Kakistocracy, and all three are linked by a reinforcing loop.
The way ahead
Blockchain with AI can lead to a world where man, while born free, will everywhere be in chains, or rather blockchains (to paraphrase Rousseau). These tamper-resistant chains will start to form upon birth, irrefutably tie us to our records and shape our future opportunities. Complex situations requiring rapid decision-making will be increasingly relegated to AI as humans in the loop will increasingly slow things down, forming a cycle where its growing abilities reinforce dependence on blockchain empowered by AI. The ever-increasing reliance on blockchain and AI will be further fostered by the virtual environments (metaverses) with their economies that social media may evolve into and the rise of digital assets such as NFTs. We may reach a point where it cannot even be said that we would have nothing to lose but our chains (to paraphrase Marx), as individual identity will be tied to the incomprehensible, indestructible complex systems that sustain livelihood and civilization itself using chains forged by blockchain and AI.
However, 2084 need not be 1984.
Like any powerful technology, blockchain possesses the innate potential for good and bad. The dire scenario outlined here is one of many possible futures. Blockchain has opened the doors to much broader applications than the ones for which it has become popularly known. It is often difficult to predict the undesirable effects of technology during its process of adoption. Furthermore, adoption is often accompanied by optimism and biases that lead champions of the technology to be blind to potential adverse effects¹³; a message disseminated widely with the adoption of that technology. An uncharitable way of putting this would be to quote Voltaire, who wrote that “it is difficult to free fools from the chains they revere.” Indeed, the followings associated with some cryptocurrencies have been compared to cults, positively or negatively, and come in for much criticism and countercriticism. Regardless of the extent to which these observations may be accurate, what is required is an inquisitive approach that, while hopeful, is open to recognizing emergent undesirable outcomes and pivoting if deemed necessary. It is essential to continue learning by trial and error¹³, especially for solutions deployed into complex social or economic systems¹¹, rather than be wedded to the favored solutions with which we start out. We must not look only for a deterministic approach to problems but also for wisdom, a holistic approach that accounts for the potential for shortfalls.
The choice is still ours to make. But complexity engendered limitations to predictability makes choice a feedback-based system rather than a binary choice.
The paucity and lacunae in the discussions about pitfalls from blockchain technology led to this article, especially in contrast to the very high degree of optimism amongst its proponents. We focused on threats to the early vision of blockchain that flow from the capabilities of the technology. While what we outlined is very plausible, we are proponents of blockchain technology. We remain optimistic about the benefits. This means that while we believe that there is a significant probability that the net results will be beneficial, there is also a significant but less probable possibility for harm. For blockchain to succeed, there must be a strong contrarian argument. Our optimism for blockchain technology at cybereum is reflected in our mission to build next-generation DAG DLT empowered by AI to improve the efficiency of human endeavors undertaken as projects. Token generation on our platform will be from real project work rather than from mining. We recognize the decentralized nature of knowledge; Our vision is based on recognizing the potential of blockchain to resolve the contradiction between centralization and decentralization by bringing benefits associated with centralization, direction, single-source-of-truth, and information dissemination, to decentralized systems¹⁴. Our platform architecture will enable benefits hitherto associated with centralization, such as responsibility, verifiability, a single-source-of-truth, and data dissemination, in an even more effective manner, without requiring centralization¹⁴. However, we recognize that while studies on project efficiency emphasize the collaboration and autonomous self-governance enabled by our architecture, the same architecture can also conversely enable centralization to a degree unattainable previously. We have outlined risks from improper use of technology that we are creating. Risks are things to be identified, recognized, and managed.
Blockchain is here to stay. Blockchain technology possesses the potential to power dystopia or utopia. The future will depend on our ability to choose well, and crucially, to learn from our choices.
[1]: Azucar, D., Marengo, D. & Settanni, M., 2018. Predicting the Big 5 personality traits from digital footprints on social media: A meta-analysis. Personality and Individual Differences, 124(1), pp. 150–159.
[2]: John, J. O., 2000. Stasi: the untold story of the East German secret police. 1st ed. Boulder, CO: Westview Press.
[3]: Müller, V. C. & Bostrom, N., 2016. Future Progress in Artificial Intelligence: A Survey of Expert Opinion. Berlin: Springer.
[4]: Kahneman, D., 2012. Thinking, Fast and Slow. London: Penguin.
[5]: Tsao, A. et al., 2018. Integrating time from experience in the lateral entorhinal cortex. Nature, 561(1), p. 57–62.
[6]: Werndl, C., 2009. What Are the New Implications of Chaos for Unpredictability?. The British Journal for the Philosophy of Science, Volume 60, p. 195–220
[7]: Hitchins, D. K., 2007. Systems Engineering: A 21st century systems methodology. 1 ed. Chichester: Wiley.
[8]: Taleb, N. N., 2007. The Black Swan: The Impact of the Highly Improbable. 1st ed. New York: Random House.
[9]: Rittel, H. W. and Webber, M. M., 1973. Dilemmas in a General Theory of Planning. Policy Sciences, Volume 4, pp. 155–169.
[10]: Hester, P. T. and Adams, K., 2014. Problems and Messes. In: Systemic Thinking. New York: Springer International Publishing, pp. 23–34.
[11]: Ronfeldt, D. and Varda, D., “The Prospects for Cyberocracy (Revisited)” (2008). Working Papers. Paper 29. http://opensiuc.lib.siu.edu/pn_wp/29
[12]: Danaher, J. The Threat of Algocracy: Reality, Resistance and Accommodation. Philos. Technol. 29, 245–268 (2016).
[13]: Taleb, N. N., 2018. Skin in the Game. 1st ed. New York City: Random House.
[14]: Natarajan, A., 2021. Distributed Ledger Architecture for Collaborative Megaproject Management with Authenticated Participants. Shanghai, ACM.
