While on an overnight bus ride in Peru, my wife and I watched The Remarkable Life of Ibelin (2024). It is a deeply moving documentary about the life of the late Mats Steen, a Norwegian man born with Duchenne muscular dystrophy who started playing World of Warcraft intensively as it became increasingly challenging for him to participate in the physical world.

Mats’ virtual life reveals a world onto itself, seemingly distinct from the “normal” physical world of his parents and siblings until they collide after the unfortunate occasion of his passing. There’s a touching story here based on the experience of Mats and his friends from World of Warcraft—about the potential humanness of our online spaces, where friendship, love and meaning can be fostered, extending beyond the boundaries of the digital.

At the same time, over and above the specific story of Mats’ life and legacy, I think the documentary viscerally highlights a more general phenomenon about the Internet: how it enables or even encourages the multiplicity of the self.

Living many lives online

In Mats’ case, the virtual world within World of Warcraft allowed him to transcend his physical limitations, which in turn gave him a new context to socialise with others outside of his immediate physical environment in Oslo. Here, Mats’ digital actions carried far more weight than his physical capabilities, and his digital identity could manifest independently from his physical one—at the very least, he had the agency to decide how to bring together his digital and physical identities.

However, it is not only online videogames with immersive virtual worlds that can empower us to cultivate and enact different versions of ourselves. The ability to do so is a superpower of the Internet more broadly.

While humans have always had the ability to harbour multiple identities, there were sheer physical limits to this. We could only ever be in a single place at a point in time, and inhabit a single body to interact with others. That said, as digital tools on the Internet evolved to become more feature-rich and accessible, transcending these constraints became easier.

Even though the consumer Internet is not even as old as a single human lifetime, anyone with an Internet connection now has the ability to “exist” in multiple virtual environments. We can embody different avatars and represent different facades of ourselves while doing so.

After all, the Internet is fundamentally multi-modal. There are a variety of ways for us to be online—a variety of platforms to use to engage with others virtually. For example, I can behave with propriety on Facebook where my parents and teachers can see my posts, while sharing dank memes and shitposting with impunity in more private channels like WhatsApp and Telegram group chats or private Instagram accounts. At the same time, I can troll on Reddit and create brain rot videos on TikTok, while polishing my professional credentials on LinkedIn and curating a pixel-perfect visual feed of my ideal life on Instagram.

We enact our identities in diverse ways within a single platform too. Although many social media platforms were founded on the premise of unifying our digital identity with our physical one in a single profile, users have challenged this by creating alternate, often pseudonymous, accounts on these platforms. Finstas or alt Twitter accounts are a case in point. There, we express ourselves in ways that we wouldn’t on our “real” accounts, or pursue more esoteric and edgier interests for which we would have difficulty finding a community “in real life”. In doing so, we inevitably construct more multi-faceted and layered social realities for ourselves—our identities manifesting across the continuum between our physical existence and virtual imagination.

In a similar vein, Ben Thompson, who is behind the tech publication Stratechery, had postulated about how social media would move on from “forcing everyone to be their ‘whole selves’ for the world” to enable “people to be themselves in all the different ways they wish to be”. This iteration of social media, which he dubbed “Social Networking 2.0”, is based on the premise that it is natural for users to fragment their selfhood online, and enact different facets of themselves based the specific context they inhabit:

What I increasingly realize, though, is that separating my identities on Twitter does not mean a lesser experience, but a far superior one; social interaction in any medium is always a balance between self-expression and the accommodation of others, which means that in the analog world it is a constant struggle to strike a balance between being myself and annoying everyone around me at some point or another. The magic of the Internet, though, is that you can be whatever you want to be […]

Expanding who we can be online

Needless to say, the Internet has dramatically expanded the possibility space of our selfhood, with our identities becoming more diffused and dynamic than ever before. And if you, like me, think all of this is weird, we’ll soon have to reckon with things becoming even weirder.

Simply extrapolating from some emerging technological trends, we’ll have even more tools at our disposal to further extend the multiplicity of the self and expand who we can be online:

(1) Digital property rights

One of the most established use cases of blockchain is as a custodian of property rights for digital objects. A non-fungible token (NFT) on a blockchain is tantamount to a certificate of ownership, one that is openly verifiable and tamper-resistant to boot.

I have little doubt that the phenomenon of owning tokenised digital objects will become more established in the future, given that ownership is a superior way of engaging with meaningful digital objects as compared to the current modality of ephemeral consumption—do you even remember which posts or stories you liked on Instagram last week?

As the literary critic Walter Benjamin had written in an essay called “Unpacking My Library” in 1931:

[O]wnership is the most intimate relationship that one can have to objects. Not that they come alive in him; it is he who lives in them.

Indeed, our sense of self is often shaped by the things we collect, our tastes defining our identities. This should apply as much to digital properties as physical ones, all the more so as our lives are becoming increasingly digitalised. In this context, owning digital properties—through tokenisation—will certainly enrich our relationship with the virtual realm, and empower us to be more intentional about stewarding our online footprint and identities

I can speak of this from personal experience, with my relatively newfound interest in collecting art on the blockchain now contributing to a significant part of my identity. I am thus not exaggerating when I say that some of the art I’ve collected have profoundly influenced my worldview, consequently expanding what I regard to be my selfhood.

(2) Durable, programmable, composable and networked digital objects

Another key use case for blockchains is as both a canvas and computer. In this regard, I believe that blockchains can facilitate the advent of a new class of digital objects imbued with many simultaneous special powers: durability, programmability, composability and networkedness.

Digital objects with such special powers dramatically open up the design space for digital and online interactions. In turn, they can provide novel avenues for us to forge meaning in the virtual realm and beyond. Some speculative thoughts in broad brushstrokes:

• Durability in the digital objects we own will compel us to consider our digital legacy more intentionally. When we can create and collect onchain digital objects that will outlive us, what we do with them will speak volumes about what we want to be remembered for.

• Programmable digital objects can be designed to dynamically evolve based on their owners’ actions or predetermined inputs. This not only gives us agency to shape our own digital experiences, but also connect us more intimately with our ever-shifting digital contexts.

• Composable digital objects can be built on by others to be remixed, recontextualised, and reimagined permissionlessly. The key here with regard to identity-formation lies in the creative lineages that such composable digital objects foster. By engaging with such digital objects while ensuring that this engagement can be immutably tracked on the blockchain, we can more effectively ground our digital footprint and the identities built on top of it in a broader cultural reality that is shared and continually evolving.

• Digital objects networked together on a blockchain can serve as loci of coordination among owners or even other users. Such objects manifest their own spheres of emergent behaviour around them, enabling the actions of a broader collective to influence our individual relationships to these objects, and consequently, our identities.

In sum, such durable, programmable, composable and networked digital objects can engender new forms of digital behaviour or prompt new ways of thinking about them, especially when their special properties converge and build on one another. As our lives become increasingly digitalised, it shouldn’t be a stretch to think that such digital objects will similarly have an outsized impact on our own identities and sense of self.

(3) AI agents

The term “artificial intelligence” (AI) has become entrenched within mainstream consciousness  ever since OpenAI opened up ChatGPT to the public in 2022, and made it accessible for the average person to experience for themselves the wide-ranging capabilities and versatility of such general-purpose large language models (LLMs).

Many other AI-assisted tools have also come into the consumer market during this time, and there currently seems to be significant excitement for so-called “agentic AI” or “AI agents”, especially when they are let loose on permissionless crypto rails. Such agents tap on the voluminous quantities of data that their underlying models were trained on to be able to autonomously perform a variety of tasks, from simply shitposting on Twitter to purportedly investing like a venture capital firm.

While much attention has been devoted to what such AI agents can do (given the prospect of profits to be made), what I find more intriguing is the question of what we can become with such AI tools.

My crude mental model of AI is that it is simply a tool to leverage on data—utilising maths to map patterns within datasets and thus open up new possibilities in terms of what this data can do. For example, we can train a model on lots of writing to create a writing/editing tool, on lots of images and text captions to create a text-to-image generator, on lots of human conversation to create a virtual companion, and so forth. Similarly, I do wonder: what if we turn the power of AI on ourselves, and use AI agents to leverage on our personal data to expand who we can be?

In this regard, how could an AI-augmented self look like?

• Scaling our skills. The most obvious use case for personalised AI agents is simply to scale up the actions we do well, trained on how we would perform those actions. In other words, these AI agents can serve as virtual extensions of ourselves, amplifying our personal skillsets in ways we could have never done before. For example, if I was a designer, my personalised AI agent could design works based on my personal style, and do so in prodigious quantities without having to take any breaks.

• Simulating selfhoods. Less obvious in my opinion is the potential of AI agents to serve as an incubator for alternate identities: to simulate different versions of ourselves and see how these different versions evolve if we varied certain parameters as a proxy for our personalities or pursued a specific set of actions or interests. Given that AI agents can already replicate the attitudes and behaviours of real people, it shouldn’t be a stretch to imagine that they can also be harnessed to experiment with and develop alternative identities. This is a natural evolution from the fragmentation of identities already happening when we bring different aspects ourselves onto the different platforms we use online. AI agents just enable us to do this more intentionally and aggressively. Instead of manually managing a couple of social media accounts today, I can now be “live” across so many more platforms and contexts through my ensemble of AI agents—each one imbued with a bit of myself while also tweaked to be slightly different. With this capability to simulate multiple selfhoods and let them play out in “rehearsal spaces” with other humans and agents, our digital lives may very well become fuller and richer—effectively n-dimensional—encompassing the manifold possibilities and trajectories that were once impossible to pursue all at once:

(4) Programmable cryptography

Today, most of the data flowing through the Internet is cryptographically-encrypted in some form. For example, a vast majority of websites now utilise the encrypted HTTPS protocol, which helps to protect the integrity of information that users send or receive when connecting to them.

In the coming decades, however, it is likely that not only data will be cryptographically-secured, but computing itself. The 0xPARC Foundation has used the term “programmable cryptography” to refer to the emerging set of technologies that will enable general-purpose computing to be performed on top of or inside cryptographic protocols. Examples of the cryptographic primitives that would fall under the umbrella of programmable cryptography include zkSNARKs, fully-homomorphic encryption (FHE) and indistinguishability obfuscation.

Understanding how these programmable cryptographic primitives work goes way beyond my level of technical competency, but what is interesting to me are the implications of programmable cryptography on our online lives. Broadly, such primitives can allow us to run computing programs with unprecedented privacy-preserving, security and interoperability guarantees: they can never learn anything from the private data they use to compute and they can also execute with cryptographically-verifiable correctness, all while remaining permissionless to use and build on.

Examples of applications of programmable cryptography, as suggested by the 0xPARC Foundation, include:

• Getting personalised recommendations for entertainment, education and professional opportunities, healthcare services, financial products, etc., without your personal data ever leaving your computer and being acquired by a third-party, even the entity providing the recommending service.

• Having a universal and interoperable set of digital identity standards that can be used seamlessly, securely and permissionlessly across all digital platforms, while your underlying personal data remains private and under your control.

• Creating self-executing, self-sovereign AI agents that can hold their own secrets and never go down, while being perfectly interoperable with other digital platforms and services.

Even though these programmable cryptographic applications certainly don’t seem tenable anytime soon, they point towards a future in which personal privacy and agency are fundamental to the online experience. In such a context, it should no longer be a norm for us to hand over our personal data to third parties to be able to use a digital service. Instead, we will each be able to determine the exact data a third party can access, without having to trust any intermediary.

Ultimately, given its privacy-preserving and security features that are enforced not by hardware or software but by maths itself, programmable cryptography will significantly bolster our confidence to bring even more of ourselves into the digital realm. And when we do so, I believe we are going to turbocharge all the other dynamics I talked about earlier regarding the use of blockchains and AI agents, and potentially create other new, emergent behaviours.

Conclusion: Squads, swarms and shields

It is clear that the Internet has opened up the playbook for how we can conceive of our own identities and selfhood, and that we already lead many different lives in the digital realm. But even as we endeavour to adapt to this reality, we should expect that our digital lives will continue to radically evolve and expand as a new class of digital technologies mature.

Although it is uncertain how technologies like blockchains, programmable cryptography and AI will eventually pan out in terms of mainstream use cases, it is nevertheless important to start updating our cognitive toolkits regarding these emerging technologies. After all, they hold immense power to reshape the very foundations of our digital identities.

In this regard, I’ve been rethinking what exactly being an individual entails in this next phase of the Internet. This essay represents the rough sketches of my current thoughts, with the following being a more concise distillation:

• Emerging technologies like blockchains, programmable cryptography and AI will further exacerbate the fragmentation of our digital identities.

• Blockchains and programmable cryptography are poised to imbue our digital world with physical properties (e.g. ownability, scarcity, and privacy), while leaning in on attributes that are natively digital (e.g. scalability, verifiability and interoperability). This opens up new possibilities in terms of what we can do on the Internet, and consequently, what we can become online.

• At the same time, AI will not only enhance our capacity to act within both the digital and physical worlds, but also empower us to simulate anything, including ourselves. In other words, with AI, we can further segment our already fragmented identities into distinct simulations, which can then autonomously take on “lives” of their own. Our selfhoods will be multi-dimensional by default.

• In this next phase of the Internet, our individual sense of self may thus be founded upon squads of agents manifesting a spectrum of simulated identities, all managing and engaging with swarms of ownable and dynamic digital objects, while being protected by programmable cryptographic shields that restore personal privacy and agency to the foundation of our online experiences.

With such squads, swarms and shields, I think the potential for living expansively online has never been greater. Notwithstanding all the pitfalls of the contemporary digital paradigm, the Internet is now an unstoppable phenomenon and it is surely in our interest to learn how to cultivate better versions of ourselves within its omnipresent embrace.

Disclaimer: Nothing in this essay constitutes investment advice. Please do your own research or consult your own advisers concerning any potential investment decision.

Credits: The header image of this essay is a crop of a post-photographic artwork titled “Impostor syndrome” by Alice Gordon, part of a larger series called “Cognitive Behaviour” that was released on Fellowship in July 2023. The full work is displayed below and it is currently in my personal digital art collection.