Immutability Hell28 Jul 2021
Among the immense promise of so-called “Web3.0” is that of a decentralized and “permanent” web, with the linchpin of either quality being something like finality. In other words would we cite the irreversibility of, say, transactions or contracts or, more generally, interactions, as well as the intractability of the information entailed as characteristic of this paradigm; implicit here is the immutability of the record of occurrence as well as consensus upon the preservation of such as the necessary criteria for the achievement of said finality. While all of this should undoubtedly prove integral to any concerted and consequential effort toward the decentralized infrastructure characteristically sought after under such a paradigm, it is also among its most intimidating when it comes to publishing challenging work or anything which might be construed as contentious, let alone anything personally revealing by means of that paradigm’s associated infrastructural ecosystem. (its network protocol or stack)
While this is a feature of great utility for ensuring posterity and resilience against censorship or any such loss of context, it may also, for good or ill, preserve in all likelihood a berth of mistakes and infractions or occurrences otherwise regrettable or damning. Again, this may be considered a feature in terms of preserving context (such as for a public grievance or exchange of any sort), but even in relative obscurity, it may come to seem as though an albatross around one’s neck in the very best of unfortunate cases, or else in the worst of cases as practically a death knell.
One might argue that denizens of the Internet at large have indeed proven in dire need of a rigorous course of discipline under the considerably less auspicious realm of permanence intrinsic to the new paradigm–a sort of social-engineering for the next epoch of digital interconnectivity. And while certainly on some levels might this school of hard knocks approach prove a necessary and even desirable shift to contend with, there are ways in which to alleviate the collateral damage of what I term immutability hell and sufficient reason to do so.
The Freedom to be Forgotten
This is a rather young and contentious concept, the so-called freedom to be forgotten1 but an important one to balance in consideration of maintaining individual autonomy upon a backdrop of persistent and pervasive transparency. In our own case, we are dealing with a matter of consensus in terms of common control over a common ledger (as opposed to public constraints over a private data hoard and the transmission of its particulars), but we should not underestimate the degree to which this consensus is come by in a passive yet, in equal measure, fickle manner; by which is meant that though the average user will care little about the particular nuances of the ruleset under which they participate, this should not be mistaken for authentic consent but, rather, a frivolous sort of trust, intrinsically precarious by the almost uniform heedlessness of its granting.2
This may sound strange or inappropriate: that users would allot trust in regard to what is so often fundamentally trustless infrastructure, but while such infrastructure operates and exists almost purely by mere consensus in practical terms, such infrastructure may live or die by the broadness of such consensus (its claim upon legitimacy3). Like all things, disintermediated and otherwise decentralized platforms and protocols alike are subject to market forces and, thus, success therein to stay aloft the oblivion of vaporware; this, at the very least, means a sufficiently active user-base, but it likely also means a reliable network on one or more levels. Without trust, marketability or, more accurately, any claim upon legitimacy suffers, regardless of any inherent or perceived level of trustlessness, which can never preclude or substitute for the desire and expectation of trustworthiness. This goes as much for adherence to standards both documented or inferred, even if some such inferences happen to contradict the documentation or intent of design.
Ironically enough, even one of the chief protocols endemically associated with the emerging paradigm of “web permanence” facilitates a hefty degree of impermanence by design: IPFS, for instance, practically automates the discarding of unsupported content, with nothing more than a unique hash to indicate any prior existence, if even any record of that hash were itself preserved. What we here encounter is a suggestion for just how to mercifully retire information which garners no attention (or else remains unsupported, under-served, or what have you) and hence serves no context. What then remains to discern is how such retirement might be facilitated without imposition.
Moderation and Iteration
Iteration, moderation, and, for that matter, self-moderation are all too often taken for granted in the intrinsically centralized commercial approach of “Web2.0,” epitomized most glaringly in the form of contemporary social media. The commercial viability of one approach over another can be as deterministic and trivial as “the customer is always right,” which usually results in the user wielding carte blanche privilege over editing and/or deletion and “damn all else.” But of course the “customer” is always some amalgamation of the userbase (itself some atomized aggregate), advertisers, and data-miners of whatever other sort, and nowhere is this commitment’s dynamics so clear as in the usual manner of such platforms’ wielding powers of moderation; this last point perhaps goes some ways to explain the apparent lack of ambivalence with which the foremost social media enterprises prosecute the matter of iteration and moderation.
Decentralized approaches, on the other hand, find little luxury in so easily abating such considerations to mere matters of profit and must ultimately reflect a general will in a way which mitigates unease as well as the sort of damage of the sort such platforms prove increasingly capable. For instance, how might one preserve context when either iteration or deletion might be levied toward its distortion? Or how might deletion be represented or even accomplished in the face of permanence (that of the swarm and/or the distributed ledger, say) if not at the expense of such context?
It would surely prove no trivial matter to resolve such in a way which does not sow confusion or violate trust. One approach, however, might be discerned by simply attempting to authenticate, in as unobstructed a manner as possible, the mess of social interactions which underlie their digital representation, rather than trying to pave them over, whether in the sense of a smoothing over or even as a veritable fossilization (or preserved as though in amber) or enshrinement of the history of such interactions. (immutability for its own sake)
“We live in a society.” (or “Hell is other people.”)
The manner in which we publish can either signal intent, or it can fail to do so. Carving an innocuous thought into a brick wall cannot but rob that thought of any such innocuousness, rendering it firmly a provocation in whatever regard. Were such an inscription to later be paved over or otherwise defaced for either the sake of re-inscription or simple redaction (or beautification, for that matter), the original provocation would either be abated (in the case that no new context had proceeded from its original inscription) or else deepened (in the case of its expunging or distorting whatever context had indeed emerged of the original). The question then most appropriately falls to the matter of dominion over that brick wall, particularly as a medium of public discourse to be managed as such if allowed to persist under whatever domain.
For whatever reason, the public (for lack of a better term) tends to inflict upon itself (or acquiesces to) certain limitations in exercising agency over whatever medium of discourse might be regarded (however erroneously) as outside its domain, despite the actuality of that discourse occurring precisely within the public domain (emergent in and of the general intellect, cultural memory, and so forth). Considering this might we imagine, let us say, a wall which is publicly owned and managed–designated for just such a purpose of inscribed discourse. How might the public manage such a thing? Who decides? Indeed might we ponder ceaselessly at what manner of arbitration be brought to bear in deciding just who is to be empowered toward the purpose of alteration and redaction and in what context, let alone any other manner of moderation which might strike the public’s fancy.
One thing seems certain: The less any resolution to such matters leans upon centralized intermediation, the more alike it might seem to an actual community of any sort–particularly one considerably less generally mediated. Let us consider a moment, for instance, just what sort of mediation occurs in the contemporary form of DLT. (Distributed Ledger Technology)
When something occurs on a typical blockchain, it is not simply permitted to occur in the sense of some vested authority or executor in effect acting as proxy in doing the deed, rather does it happen because it can happen–because it meets the criteria of having happened. Now clearly any such criteria, in order to be met, must needs make reference of the ledger hitherto and must itself be committed to that selfsame ledger in order to bear consequence. (i.e., to be referenced as having happened)
Interestingly, transactions facilitated by a blockchain and its respective network, take on many of the characteristics of a cash transaction with the innovations upon that manner of transaction including worldwide permeation and the lack of distinction between a transaction and its receipt. Nevertheless, the lack of any need for an intermediary means that such transactions bear similar counterparty risk, owed to their fundamental intractability, meaning that any discrepancies must be handled rather personally–a circumstance seldom seen among other manners of electronic payment. Yet, it is true that even personal disputes usually occur and resolve (or fail to do so) under communal context of some sort, and that context always augments such disputes–renders them more than personal–with each party in their own position beholden firstly to the community if at all to any entity.
Community can pose a hell unto itself, however, and so must it fall to any given community to mediate and mitigate its own hell that it might persist and thrive (thriving as itself a condition of persisting) as a community. Whether etched centrally as into brick, or inscribed widely into hot silicon, a community must find a way to persist in spite of its permanence–in spite of its perpetual casting of its own death masks–if that community is to remain living through its immortalization, rather than perish of it.
This is perhaps the very crux of our matter and why it would behoove any such community as might emerge under this new paradigm to contend seriously with the encroachment of immutability hell for its members that it might stave off its own ensuant petrifaction in the process.
Relating here generally to the underlying concept and implications of the “Right to be Forgotten” as it appears in the GDPR, though we do not here deal implicitly with the law or any specifically legal matters of inquiry. ↩
Relevant to this is a particular insight from a piece regarding the Byzantine Generalization Problem: “Blockchain technology claims to be trustless, in the sense that individuals don’t have to trust intermediaries or more powerful single actors in order to act and interact. The technical system off-loads authority onto a transparent and public consensus history, created and validated by the protocol and some of its users. The technical system could be considered a “trustless,” multiauthored actor in its own right, with the however unlikely and expensive scenario of its own validators’ collusion. “We don’t need to trust each other before we can begin to collaborate,” blockchain technology claims. Through the massively poor media reporting on blockchain, this claim can come to be misinterpreted, as today’s dominant connotation of trust suggests something interpersonal and chosen. The term “trustless” skims over that what may have been previously defined as “trust”—as in trusted institutions—and in many scenarios may not refer to voluntary, cultivated relationships but instead arise as a result of contingency and lack of alternatives. In this light, trust was the network effect of rumors around consolidation of power.” [emphasis added] ↩