Self Actioning System, a preferred systematic embodiment of “Digital Self”
Part 1: Digital Twins in a Surveillance Economy
This article is the first in a two-part series. It aims to clarify the benefits and limitations of digital twinning technologies and why recent developments in AI continue to raise global concern regarding the inevitable socio-ethical dangers of digital twinning conscious entities–––a virtual model of "digital self"–––and our quest to protect free will and human agency in the current surveillance economy, where humans hope to leverage insights from digital twins for material objects and processes rather than being beholden to a virtual representation of "self" in the digital space.
Waking up to an unwelcome reality: Before the Internet of Things (IoT) brought the term "wearables" into fashion, people have been extending our physical and cognitive abilities for generations through supportive inventions such as glasses, watches, walking sticks, hearing devices, and wheelchairs, or used mind-altering consumables to extend our natural capabilities beyond previously known limits to assist us in our daily routines and activities, influencing the imagination of "self" as we evolve as an intelligent species.
In light of recent advancements in generative artificial intelligence (GenAI), creating a virtual representation of ourselves for self-determined interactions in the digital world is a societal issue as much as a technological one, pointing directly to our species' sense of being. Therefore, delineating between "what is human" and "what is a machine" lies at the heart of technological innovation in determining how humans engage with digital systems, with "human consciousness" a dominant factor in that delineation.
““It is a habit of mankind to entrust to careless hope what they long for and to use sovereign reason to thrust aside what they do not fancy.””
The philosophical concept of "Sovereign Reason" states that all humans possess the inherent right to have supreme or ultimate power of a self-determined cause, explanation, or justification for an action or event, acting as the sole instrument of "free will."
Analogous to marine life struggling to survive beyond aquatic ecosystems, engaging in a digital world remains overwhelming to humans. Our ability to adapt to and comprehend the ramifications of deploying next-generation technologies is still embryonic. At Human Colossus, implementing any disruptive technology without fully understanding its influence on societal well-being is unfathomable. When AI is intrusive enough to result in coining the term "Digital Twin," societal vigilance is necessary for any technology deployed under that banner. Humans are conscious entities, not material objects. Developing a digital twin of a conscious entity is dangerous. An accurate virtual representation of "self" must never be in the form of a digital twin.
Refactoring AI innovation could narrow the current analogue/digital divide, providing a more realistic digital representation of real-world interactions with other entities in a virtual environment.
Why the notion of a digital twin of a human being continues to raise global concern
A "digital twin" is a virtual model designed to accurately reflect a physical object or process. They are created by collecting enough data to model an object in the digital space with enough accuracy to serve a specific purpose.
The benefits of digital twinning material objects and processes
A simple example of a "digital twin" is a digital image captured by a digital camera viewed on the camera's image display. The captured image is the digital counterpart of a negative in a roll of film, a photographic image stored as a single frame captured by an analogue camera. By creating a digital representation of the analogue photo, the photographer can enhance the image through digital image processing, a set of underlying algorithms to process existing data about the digital photograph mathematically. The image processing algorithms may include low-level methods, noise removal, and colour enhancement. AI algorithms provide data actioning recipes to accurately process data packets to represent material objects and processes in the digital domain. You can experience the digital counterpart of a material object or process remotely.
One of the most well-known companies using digital twins is Tesla, which creates digital simulations for every car it sells. Each vehicle has sensors and sends data to the cloud, and Tesla then uses AI to analyse their performance.
Organisations use digital twins for compound use cases. By collecting enough data about a physical object or process, digital twinning enables their corresponding representation in a digital environment, where the supporting data is ingested into virtual simulations to observe performance, check for physical limitations (e.g., crash tests), stress test critical components, and verify behavioural response in varying conditions. These simulations significantly reduce operational costs. After all, how often can an aviation company such as The Boeing Company, for example, afford to crash its Boeing 737s to test their handling in harsh weather conditions. Mechanical engineers can repeatedly crash a plane in a simulation without significant financial burdens or consequences. So, digital twins are disruptive to product lifecycle management (PLM) processes in most industries; they enable product owners to access an enormous amount of real-time data collected using sensors to ingest factual information about a physical object (the "product") or behavioural information about how that object responds to changes in environmental circumstance through its digital counterpart in simulation.
Previous technological limitations confined product owners to only being able to test early-stage life cycles in a lab before releasing a product to market. However, due to the ever-decreasing costs of digital sensors, the Internet of Things (IoT) revolution has scaled digital twinning technologies, with digital twins now able to provide algorithmic reasoning for performance monitoring at every stage of a product life cycle, with data insights about consumer products accessible in real-time. Product owners can create digital twins for a product and create digital twins for an instance of a product. That level of granularity makes past impossibilities in trend analysis now possible, with product insights more accessible across the board.
So, how can we use digital twins? Well, digital twins can run practical simulations to study multiple real-time processes, such as predicting failure, improving response times, and informing customers about maintenance information to prolong a product's functioning life. With simple controlling mechanisms built into the devices, we can monitor responses, inform product customers, and remotely solve issues before they happen, enabling automated self-healing functionality. Digital twins can fully control a device without unnecessarily burdening the consumer, allowing a more seamless user experience.
Digital twins allow product owners to monitor granular processes across an entire product lifecycle, including at all stages of research and development (R&D), design, manufacturing, service and maintenance, and daily operation.
The socio-ethical dangers of digital twinning conscious entities
For the most part, developing digital twins for material objects ("products") and processes can significantly benefit organisations wishing to streamline product development and optimise costs, and their customers, who benefit from greater product efficiency. However, if virtual representations of objects can profoundly benefit society, can we produce a digital twin for a conscious entity, such as an intelligent animal or human? A virtual model of "digital self."[1]
In truth, through corporate surveillance, many companies already have enough data about customer buying behaviour and other traceable patterns to develop virtual representations of consumers (without consent) to benefit the organisation in any digital environments they control, an unwelcome economic reality and, in light of digitally twinning civilians, a sobering thought.
For large swathes of the global population, humans are well connected as an ever-evolving species reliant on technology. However, through our various online interactions with service providers and product owners, organisations can leverage personal data gathered through their collection processes to benefit their companies. With the sheer amount of personal data collected, by integrating a small amount of consumer behavioural data, companies can run powerful algorithms to process sensitive information about us, directly profiling our deepest desires, no matter how private the information, including our watching, buying, and eating preferences, and this epidemic doesn't stop with probabilistic predictions, as simulated models can coercively tell us what to do without relying on the intricacies of human consciousness.
Digital twinning conscious entities would inevitably leave human reasoning to machines and the ultimate loss of our God-given right to "free will."
But Free Will will prevail.
Luckily, there is an inherent flaw in the logic of profiling using big data. Although relatively simple to explain, exposing this weakness brings awareness of the risks we face in the current digital economy and gives rise to an alternative solution to digitally twinning civilians.
Even with tech companies collecting an enormous amount of data about us, there remains an unquenchable thirst for them to collect more and more personal data to fuel their algorithms. This frenzy for data is easy to explain.
So, imagine a scenario where you purchase an appliance, for example. Shortly after buying the product, you'll often receive targeted ads from a different store for a similar product at a lower price. So why do digital companies rely on this type of ad campaign? Despite having access to massive data sets and analytical tooling to generate "digital twins," the business model is flawed. If common sense prevailed, declaring that you've already bought the desired appliance should deter any marketing algorithms from sending you targeted ads for similar products after the purchase event. So, why do companies engage in this type of business model? The simple answer is “cost."
Correctly identifying helpful information in a database to enable a simple decision process is a complex and largely unsolved problem. With more and more data required for companies to maintain a competitive edge, marketing analytics resort to brute-force correlations, pressing for a tsunami of data that fuels the digital economy, leading to the societal consequences we see today.
At Human Colossus, we think differently and advocate for a Dynamic Data Economy (DDE) where the integrity of data objects and processes and the authenticity of recorded events in associated key event logs provide the cryptographic assurance and data quality necessary for accurate analytics and improved business insights (see Blog post from November 2022). A way of visualising this difference is via the Economics of Correlators.
The Economics of Correlators
In September 2021, the Human Colossus Foundation hosted a webinar titled "Core public utility technologies for a Next Generation Internet." In the context of systems with weak privacy, Dr. Sam Smith presented the concept of the economics of correlators (see Vimeo recording @19m20sec). The correlator is an entity that tries to gather data about you and correlate it with specific behaviour to create a prediction. This correlation is the basis of digital transformation in the advertising industry. In less than a generation, the sector's primary focus evolved from "know-your-customer" to "know-your-data-and-correlate," with collected consumer data enabling companies to propose their "best" offers and products to consumers based on company-controlled algorithms. Unfortunately, the previous example of the purchased appliance demonstrates that this only works as intended occasionally due to the limitations of simple economics where accurate correlation has a cost.
Gathering real-time data with proper context and accuracy is an ongoing data curation problem for humans today, with technological and economic barriers. This article will ignore governance barriers (e.g., privacy), focusing purely on technology and economics. Today's network architecture may allow sufficient data flow for simple IoT devices or physical objects by measuring a few standardised data points. However, it cannot cope with the nuanced complexity of interactions involving humans in an even more complex ecosystem. Monitoring all that at scale is a challenging technological task with a cost. Therefore the economic constraints of extracting real-time data outweigh the benefits payoff for those companies.
The economics of the correlators describes a relationship between correlation costs and value from which we can identify the domains in which there is an extractable value in correlation. As we can see in the diagram below, the correlation costs need to be lower than the correlation value, so extracting the value from the data would be profitable. By understanding the dynamics of these curves, one can control the economics around the data, influencing the behaviour of the data market.
Fig. 1: The Economics of Correlators (slight variation of original version)
The example we provided above indicates that commercially, the most valuable pieces of information are near "Now." So the services can adapt to your current needs. But, of course, it only helps if you get a proposition of the product you have yet to buy. But as we can see on the diagram, the cost of extraction grows significantly as soon as it approaches "Now." It, therefore, reduces the domain of value extraction from correlated data. As a result, this correlation cost translates into a limit of accuracy for the algorithms attempting to predict individual behaviour.
The limits of correlation
This limit is good news when considered from a higher societal perspective. Collecting data in real-time is particularly challenging and, in many cases, leads to total surveillance of a person and their community. Whether we like it or not, information is generated by us directly or captured through our interactions on social platforms, online buying behaviour, and via our friends and family through shared personal photos on social media, a personal phone number shared through a specific instant messaging app, etc. Based on collected data, companies can make robust predictions, such as inferring a social profile of a not yet-born child based on the information their mother posts on a social platform, the country in which they are living, areas, parents' background, and more.
The success of the current digital economy is primarily due to the advertising model mentioned above, with systemic flaws for everyone to observe, where user-generated data aggregated by organisations fuel machine learning algorithms that model behavioural patterns to predict our desires and shape our opinions, enslaving consumers to corporate-driven incentives, a surveillance system.
In Shoshana Zuboffin's book "The Age of Surveillance Capitalism," the author describes "Capitalism" as "a profit-making incentive that arose as advertising companies, led by Google's AdWords, saw the possibilities of using personal data to target consumers more precisely.", an observation indicating that the current label "Digital Twins" is a misnomer. Perhaps "Digital Puppets" would be more appropriate! However, as digital twins are a reality, what can be done to reverse the trend of our current surveillance society? The wake-up call highlighted above shows that, although de facto, digitally twinning human beings is a bad idea due to known privacy risks.
Actualising digital twins in the natural world
So, the digital twinning of material objects and processes can provide enough data for humans to understand how to react to and control devices to operate in a particular manner. However, if we were to apply the same principle to living beings, we would have to digitalise the differentiating factors that separate humans from machines, the essential characteristics of living beings, namely, the concepts of "Free Will" and "Sovereign Reason" introduced at the beginning of this article. In addition, we can also highlight the economic dimension, where the human instinct for financial success is the modern equivalent of survival.
Like any software tool, digital twins can assist both good and bad practices. However, influenced and determined by the community, "Good" and "Bad" are contextual. The trickier question is how to ensure that developing the right digital twinning technology will maximise the "good" and minimise the "bad." In particular, can we provide access to real-time data that makes economic sense without losing control or compromising privacy?
And our answer is, "Yes, we can!" (and we are working on it.)
Step 1: Acting on the economy of correlators as the agent of change
As we learned in the "The Economics of Correlators" section, we can manipulate how data reaches the market, influencing the costs of the correlation or the correlation value so that value extraction can happen in desired conditions through de-correlation techniques developed for privacy preservation purposes, for example. The cost of value extraction for near real-time data correlation is high for market actors but quasi-nil for the data source. Therefore, we could change assumptions about who in that situation is a correlator and drive the correlation costs to the minimum for the system. If the user collects the data themselves, costs drop dramatically since they already have devices able to collect data in real-time, with the user deciding what data they generate, record, and share. We not only make it affordable because of the scale (i.e., just one individual instead of millions), but we also change the control over the data.
Sharing for an explicit purpose will provide a high incentive for adoption and consent.
This topic becomes critically important in healthcare, where patient data is of high value to many stakeholders. Suppose a patient or an organisation trusted by the patient has complete control of shared data for a specific purpose, not inadvertently used for something else. This scenario is a game changer, facilitating dramatically improved access to purposeful data.
Step 2: Restore Sovereignty
Individuals must operate the proper tools in a secure environment to transform the end user into the correlator. With this prerequisite, they will perform such activities without worrying about being surveilled or losing sensitive information. For this to happen, individuals must regain confidence in their digital interactions.
The Human Colossus Foundation continues to shape the requirements of a SELF Actioning System (SAS), a core technological piece for human-centric digital engagement in a new Dynamic Data Economy (DDE), in HCF’s SELF Actioning System (SAS) Working Group.
That concludes Part 1 of this two-part series. Part 2 will describe how a SELF actioning system provides a superior solution for embodying "digital self" to facilitate human-led data actions in the digital realm, where sovereign reason and consciousness reside with the system's owner (a "human"), not with a virtual representation of that person (a "machine").
Stay tuned!
Dear Readers,
Maintaining a blog requires resources, including research materials, hosting fees, and the commitment of our passionate team to pioneering a next-generation Dynamic Data Economy. We rely on the support of generous individuals like you to keep our blog running and growing.
If you found this blog post helpful, informative, or inspiring, we kindly ask for your support through a donation. Your contribution will go a long way in enabling us to continue providing insightful content and expanding our reach to even more readers.
To donate, simply visit our website's donation page or click on the button below Every donation is deeply appreciated and will directly contribute to the sustainability and improvement of our blog.
Thank you for being part of our community, and thank you for considering supporting our work. Together, we can make a difference in the lives of many.
Warm regards,
The Human Colossus Foundation
