Hi Sheldrake, thanks for the welcome. I really enjoyed reading your piece on generative identity over Christmas. The underlying approach, systems/ecology, is one I am familiar with, and I liked seeing how you applied it to identity.

I’ve been thinking about identity in an on-chain context as a precursor for a wide variety of use-cases for a while now. Your piece was very useful and provided a lot of background context I was unaware of. I like pretty much all your framing, particularly how you see generative identity as a process rather than something that is static. Like resilience, identity isn’t something we ‘are’, it’s something we constantly ‘do’ – it’s a process and should be managed as such. Our conceptions of ourselves, and other’s conceptions of us, are constantly evolving, and any truly useful identity solution must be able to do the same.

Bureaucratised identity doesn’t really reflect this, but it does have limited adaptability and is quite limited in scope. Digitalising identity will have a lot more granularity, but it also opens more problems.

I want to apply a frame that I have been using in other contexts - ‘the problem of measurement’ – and see whether it can bring value to this discussion. Generally, there exists value which cannot easily be measured or quantified, and the inability to account for this value results in its exclusion from consideration unless specific attention is paid to it.

A broad problem that exists on-chain, as well as in the real world, is that we are cognitively, culturally, and computationally biased to focus on things/what we can quantify (finance, infrastructure, bits, particles, people, quantifiable attributes etc), and rarely account for the intangible or things that are hard to measure/quantify (such as social and human capital - relationships between systems/people, health, capacities etc).

This is related to the ‘what you see is all there is’ phenomenon that Daniel Kahneman describes. Or, to quote Stiglitz (2018, p.13), “What we measure affects what we do. If we measure the wrong thing, we will do the wrong thing. If we don’t measure something, it becomes neglected, as if the problem didn’t exist.”

Generally, this means that if we primarily base decisions/systems on quantifiable evidence, and half the evidence is intangible and unaccounted for, then we are basing the decision off incomplete evidence which logically entails failure of some kind.

Bringing this back to identity, there are a lot of important factors associated with identity which will not be easily quantified. So much of what and who we are is not reduceable to numbers. Eg, how do we represent relationships between people? Or their informal skills/capacities? How is it defined or represented digitally? Who gets to define the relationship/skills as they evolve? If we follow the same approach to this as we have historically, we’ll end up ignoring many of these factors which will entail significant predictable and emergent problems. But if we recognise this problem early on, and create spaces where these types of dynamics can be represented, analysed and adapted around, then we can perhaps avoid repeating many of the same problems we have elsewhere.

This is basically a definition of the problem, and how it applies to this context. I think it gels well with your idea of generative identity, in that both require constant adaptation/evolution. Does this frame make sense from your perspective?

I think looking at it this way leads to discussions about power relationships, who gets to define what about who etc. It seems to me that a core conclusion of all this is that any identity system which reaches its full potential must have balanced power relationships alongside the capacity to adapt and evolve over time and in different context.

Thanks for linking your latest piece, Sheldrake, I enjoyed this one as well. Broadly speaking, we’re on the same page but I’ll add a few thoughts/questions at the end. I also read and enjoyed your piece Collin, it’s great to see these ideas being taken seriously by a project. I’ve skimmed the holochain docs a few times and really like what you are trying to do. My interests/approach lies at the intersection of theory and practice, so it’s fitting to find one person talking theory, and the other pratice, here.

Yes, I am dwither from SCRF, though I haven’t engaged much there recently. To add a few more clues and context, I recently completed a PhD at the intersection of resilience and governance. I’ve spent much of my life interested in human systems, and I’ve spent the last 6-7 years thinking very deeply on what it would take to build a resilient society – one that maintains values and persists long term through many types of external shocks. Identity is a small, but foundational, part of that, and this is the lens/purpose I’m approaching it with.

I am particularly interested in concepts that are transferable between contexts, as so much about resilience/ecological processes is contextual. The problem of measurement, as I described it above, is one of the ways I conceptualise the consistent problems we have in governance. It also goes beyond purely human systems.

The concept can be analysed with more depth by relating it to the observer effect. The double slit experiment is relatively famous in physics where it’s demonstrated that light’s – and most object’s - behaviour changes when an observer is present. When it is measured. This is also at the core of how modern institutions/governance systems work, we measure stuff, and use that to influence the behaviour of people/ecosystems. Much of modern science is figuring out new ways to observe stuff, or new frames of reference to privilege.

I’m emphasising this as I’m trying to establish that the idea has wide applicability and seems to be something of a constant. Measuring/observing stuff changes how it works/acts. Importantly, that’s not to argue that measuring stuff is always good or bad, it’s context dependant. There are times when we want to measure stuff (i.e. for transparency) and times when we don’t (i.e. for privacy). I argue that this is a key leverage point for governing human systems, and it has significant value in an identity context as well. This can be reduced to two key questions (which are really the same question):

1. What do we see?

2. Which frame of reference do we privilege?

The aspect of blockchains that I am most fascinated by is their capacity to systematically make things visible (or invisible, with ZKPs) – where the same rules apply to everyone with no special consideration. The ability to read the chain and be convinced of its credible neutrality enables and constrains certain types of behaviours. Currently blockchains only really deal with the first question (what do we see?) and are mostly limited to tokens/finance. However, I am fascinated by the potential for this to reshape human interaction through the expansion of the first question (what do we not see?) and the introduction of the second question through a social/identity layer. I am also terrified by how this could be used to control populations if enacted in bad faith or without due consideration.

Identity, and how it is viewed/utilised, is one of the foundation pieces that will drive broader societal outcomes. The purpose of bringing up measurement/the observer effect is to argue that a key factor in determining positive or negative outcomes will be around systematising who has the capacity to choose what is and is not observable. Essentially, we are talking about structuring/dictating power relationships here.

Both privacy and transparency are fundamentally important in different contexts, and the people affected by the decisions should have a stake or representation when they are made. A system like this could be constructed relatively easily using a set of contracts – the key will be finding the right balance. This reproduces aspects of democracy, but with much greater granularity. Which can be both a blessing and a curse.

Identity, at its core, is all about a set of relationships. Everything is relational. We can split the function of identity into two parts.

1. What others (people, organisations, institutions) know and say about us a. Formally, we have this in traditional bureaucratic identity – i.e. what institutions/the state knows/guarantees about us. Drivers license, passport, qualifications etc. b. Informally, person to person reputation is very fragmented these days. It used to mean a lot more than it does now. There’s a lot of potential to make reviews, word of mouth referrals and general reputation more meaningful with better identity structures.

2. What we know and say about ourselves, other people, organisations, and institutions a. Formally, we do this for institutions and government through democracy and the like. b. For ourselves and others it’s all informal. SSI wants to formalise a lot of this, with all the complications it entails. I’d argue that a primary use case for web3 is that it allows us to structure the visibility – and who gets to decide the visibility – of these aspects with a scale, efficiency and granularity that is unprecedented. This can be leveraged for much good, or bad. I’m hoping that this framing can help us ensure the former and avoid the latter. Combined in the right way, these pieces should serve as a solid foundation for a system which can adapt and evolve much more effectively than our society currently does.

To end, I’ll make a few points and ask a few questions.

1. The primary argument is that for the purposes of governance, we can reduce much of the complexity of ecology/resilience down to observation and prioritisation of a frame of reference. What do we see? What frame of reference do we privilege? Systematising these two factors should allow us to make better, more informed, decisions at scale. a. Measurement is too constrained of a word, and visibility, sight, and observation aren’t perfect either. Awareness perhaps? None of these words really grab the true essence, eidos, of what I am trying to convey.

2. I’m aware than reductionism is problematic, which is why I’m reducing to a principle/process, rather than a set of objective or linear facts. A lot of this came from thinking through the question “what (from a resilience/ecology pov) is transferable between contexts? This is the answer I came too. There remains a lot of very important nuance which hasn’t been fully captured above, as this is a brief outline rather than comprehensive.

3. All aspects of identity are important. This includes traditional, SSI, and generative conceptualisations of identity. Just as in science both subjective and objective knowledge is important, so too are all these aspects of identity important. It’s not about figuring out which aspect is most important to generally prioritise, rather which type fits which context, and then ensuring the appropriate tools are used at the right time. a. We still need an external authority to confirm individual identity/sybil resistance. There are times when you want SSI type control over your data. And there are contexts where generative identity is essential.

4. Definitionally, it’s fair to say that identity will be different things in different contexts. There will never be just one unilateral definition that covers all of it. Rather, we need a set of definitions which are useful in different contexts, and people can choose which tool applies best in which situation.

5. What is the difference between identity and reputation? In my mind these appear to be two sides of the same coin. How can we separate who we are, from what other people think about us? The two are fundamentally related. Our conceptualisations of ourselves are influenced by how others conceptualise us. And vice versa. These assessments can be both recursive and circular. a. Figuring out how to appropriately structure this will not be easy, but it’ll probably be simple. b. We’ll also need mechanisms to distinguish between good faith and bad faith approaches to influencing the reputation of another person/organisation. But that’s a conversation for another day.

6. Sheldrake, I agree with your vision of the failure condition for web3. Can be conceptualised in my two questions. What do we see? What frame of reference to we privilege? We currently do not see the problems illuminated by social science/ecology, because we privilege the computer science/tech frame of reference.

7. I really like your reference to Nora Bateson at the end. What creates a society that maintains values and persists long term through many different kinds of challenges will not be a static system with objectively defined rules. Rather, it’ll be the maintenance of a process which adapts and evolves as the contexts around it shifts.

8. I really did enjoy your piece Collin, going from theory to practice is difficult. It’s quite easy for people (like me or Sheldrake) to spout theory, but bringing it together and operationalising it is an entirely different process. It’s where the tradeoffs become real and imminent.

9. Collin, I have a question for you which is slightly off topic. I’ve skimmed the holochain docs a few times, and really like the intention behind what you are doing. But why is this a completely separate chain? Why isn’t this a L2/L3 rollup on eth/cosmos? I think economic security will be a big challenge here (this is not my area of expertise, so maybe there’s a simple answer to this).

So, does this make sense to the both of you? Questions? Thoughts? This is most of the theory side of what I’ve been thinking on, as it applies to identity. All of this leads to some conclusions about how to operationalise this approach, and these thoughts are in the process of congealing into something solid in my mind.