Open work
altme.org
altme.org is a separate initiative on a person’s rights over their own memory: that it is held as theirs, portable between applications, granted to an assistant by consent and withdrawn the same way.
- Subject
- A person’s rights over their own memory
- Proposal
- An open protocol
- Implementation
- Not described here
- Relation to PrepGraph
- A separate initiative
altme.org
A separate initiative. This page states its intent and the obligations it proposes, and describes no system.
The argument, in four parts
Why a person’s memory should be theirs
Every assistant you use forms some picture of you. It has to; it cannot be useful otherwise. What you can establish from where you sit is less than what it holds: that something is retained, that you were not asked for it, and that you cannot take it with you.
Three things are wrong with that, and none is repaired by making the picture better. It is thin, because what one application has reason to learn about you is a small part of what is true about you. It is silent, because at no point did anyone ask. And it is not yours, because you can neither see it, nor move it, nor end it.
The asymmetry is worth stating exactly. You are the only party present in every one of these conversations and the only party who carries nothing across them. Each one begins again, and the cost of beginning again is paid by you, in repetition.
A picture a person cannot inspect cannot be corrected by that person, and a wrong inference does not merely persist, it compounds, because what comes later is built on what came before. So the argument here is not primarily about privacy. It is about who is permitted to be the continuous thing, and who is permitted to correct the record.
The demands set out below apply to every application, PrepGraph’s own included: learners can already see and erase what PrepGraph’s tutor remembers about them.
- The picture is thin
- What one application learns about you is shaped by what that application is for. Nothing joins the fragments, so no assistant meets the whole person, and the person is the only one who could join them. A test: ask an assistant something that depends on what you explained to a different one last month.
- Nothing was ever granted
- Context accumulates without a moment at which anyone asked and you agreed. Consent should be an event, not a condition of signing up. A test: name the moment at which you agreed. If there was no moment, nothing was granted, whatever the terms of service record.
- You cannot correct it
- You cannot see what is held, so you cannot say that it is wrong. Errors about you are kept where you cannot reach them. A test: ask what happens to a decision already taken on the strength of a record you were never shown.
What is being proposed
An open protocol for a memory that belongs to the person rather than to an application, and for what an assistant should have to observe in order to be let near one.
It standardises one thing: the terms on which an agent is granted access to a memory belonging to somebody else, namely the person. It does not standardise what a memory is for, what an assistant ought to conclude from one, or what makes one good. A protocol that answered those would be answering questions it has no standing to answer.
Existing open work has settled how an agent reaches context that belongs to a service. The open question is how an agent reaches context that belongs to the person, on that person’s terms and revocably. The protocol is about the second question and takes no position on the first.
What follows are the obligations it places on whoever holds a person’s memory. They are written as demands a person, or a school buying on a child’s behalf, could put to any supplier and expect a straight answer to. How such obligations are met is not described here.
- The memory is the person’s
- Not an account record held on a person’s behalf by whoever is currently serving them, but theirs, with the application in the position of a guest. A guarantee in a document is only as strong as the implementations that keep it, which is why the protocol is open.
- Access is scoped
- A person should be able to let an assistant have some of what is held about them, for a purpose stated in advance, rather than all of it as the price of using anything. A test: ask a supplier what their application can still do when it is given less. If the answer is nothing, nothing was scoped.
- Access is revocable
- A person should be able to end a grant they made, and ending it should be an ordinary act rather than an exceptional one. The hard part: what has been read has been read. Revocation after the fact is the open problem every design of this kind has to face.
- The memory is portable
- A person should be able to take their memory out and put it somewhere else, including into a system built by a competitor, without the incumbent having a say. A test: an export that can only be re-imported into the system that produced it is not portability.
- The person can see it
- What is held about a person should be visible to that person, in terms they can read. Legibility is part of the obligation, not a courtesy on top of it. A record disclosed in a form nobody can read satisfies the letter and defeats the point.
Why this is published openly when the rest is not
This site keeps its methods private, yet it argues for an open protocol. The two are different kinds of thing, and the difference is the point of this section.
A protocol is an agreement, and an agreement with one party to it is not an agreement. The obligations above become obligations only when anyone can implement them, and a person can move between two implementations without asking either.
How any implementation works stays private.
Our own interest is plain. We would rather build on an interface we do not own than own one that nobody else uses.
What it has to do with learning
A learner’s context is the clearest case of something that should not sit inside one vendor. What a student has understood, where they stalled, the misconception they have carried for two years, the language they think in before answering in another one: today that belongs to whichever platform happened to observe it.
What moves with an Indian student is the mark. Marks transfer between schools, boards and tutors; they are the one portable artefact in the system, and they are a late scalar that cannot say what a student should do next. A teacher who inherits a section of fifty-odd in April receives last year’s marks and almost nothing about how they came about. The compressed summary travels and the understanding does not. That is backwards.
A learner should be able to carry their own context across a change of teacher, platform, medium or school. That is a position, and this programme sets out to test it.
- The question
- Does a learner who carries their own context across a transition get met at the right level sooner than a learner who begins again? The question is asked about one learner across one change of teacher or school, not about a cohort and not about a platform. Stated as a question, with the finding that would settle it below.
- What would falsify it
- If transitions go no better with carried context than without, portability is a case about a person’s rights over what is held about them and not a case about learning, and it should then be argued only on those grounds. That would not make the protocol wrong. It would make one of the arguments for it wrong, which is a different thing.
- Who holds the grant
- For a school-age child the grant is exercised by an adult, usually a parent or a school. A memory that is the person’s is, for a minor, administered by somebody else, and the person it describes is the one party in the room without a say in it. This is the hardest case for any design: it must neither give a child a power she cannot yet use nor give an institution a power over a child it should not have.
- The risk any design must answer
- A record that follows a child through school can harden into a label, and a portable memory is also a portable dossier. Visibility, scope and revocation reduce this and do not close it. A school should be able to ask any supplier what a child’s record says about them in five years’ time, and who can still read it then.
The test
A standard that only its author can implement is not a standard. It is a product with a specification attached.
That is why the protocol is open. Why this is published openly when the rest is not.
altme.org
A separate initiative
altme.org is a separate initiative. This page states its intent and the guarantees it proposes, and describes no system.
altme.org
Not a PrepGraph product, and not an offer of anything.
What a useful comment is
The most useful response is one that says a guarantee is stated wrongly, or that no honest implementation could keep it.
Continue
The same question, asked about a learner
The obligations above are stated on behalf of a person. What they mean for a learner is the subject of the memory page.