I start with the person.
That sounds obvious until the systems around identity enter the picture. Then the person gets turned into an account, an identifier, a customer record, a risk score, a government-issued number, a profile, a login, a wallet address, or some combination of all of the above. Each can be useful. None is the whole thing.
The Austrian school of economics has a useful starting point here: value is subjective, and human action begins with individuals making choices under their own conditions. The individual is not a rounding error in a macro model. That part lands for me. The same instinct carries into identity.
That does not mean every identity system has to become a heroic anti-state statement or a permanent war against every platform. It means the default should be clear: the person has meaningful control, can understand what is being shared, can revoke access where possible, and is not quietly trapped because one institution decided the account is the person.
More than a wallet
On-chain identity is often discussed as if attaching a profile to an address solves the hard part. It does not. An address can help with ownership, history, permissions and coordination. It can also be pseudonymous. It can be lost. It can be compromised. It can reveal more context than someone intended.
A wallet is a tool. A person is not a tool.
Identity involves context, relationships, reputation, consent, recovery and the ability to move through different parts of life without handing one party the master key to everything. If the only recovery plan is “do not lose the keys,” we have built a useful primitive, not a complete identity system.
The redundant phrase
“Self-sovereign identity” is a slightly funny phrase. Identity should be self-sovereign by default. Saying otherwise is a bit like saying s*tan*c (censored it again, aha!) black metal. We already got the memo.
The phrase exists because the current default is often the opposite. Identity is held together by institutions, platforms and intermediaries. They issue it, store it, observe it and sometimes revoke access to the parts of life attached to it. That can be convenient. It can also be a very concentrated kind of power.
The point is not that all intermediaries are evil, or that every person wants to personally administer every credential forever. The point is that convenience should be a choice, not a quiet transfer of ownership.
Control is not isolation
There is a bad version of this idea where “sovereignty” becomes total isolation: no help, no recovery, no delegation, no trust, no social layer. That does not describe how people actually live.
Better identity systems should make trust visible and adjustable. They should let a person delegate a permission without giving away everything. They should make recovery possible without turning recovery into somebody else’s custody. They should be legible enough that normal people can tell what they are agreeing to.
That is harder than putting a profile picture on-chain. Which is exactly why it is more interesting.
What I look for
When someone makes an identity claim, I keep coming back to a few unglamorous questions:
Who controls the identifier? Who can read the data? Who can change it? Who can revoke access? What happens when a key is lost? Can the person leave without losing their history? And what part of the system is actually theirs, rather than merely displayed to them?
Those questions are not anti-technology. They are the difference between a neat demo and something that might hold up around actual people.
Not a macro problem first
Identity is often discussed at the level of national infrastructure, institutional efficiency or global coordination. Those questions matter. But I would rather start smaller: the individual person, their agency, their risk, their ability to say yes, no, not now, or I am leaving.
Start there and the bigger system has a better chance of being worth the trouble.
◆