Herman Melville wrote a story about a curious fellow, a scrivener: a clerk for his employer upon Wall Street, a “conveyancer and title hunter, and drawer-up of recondite documents”.[1]
Bartleby’s disinclination to do any work was, repeatedly and almost without variation, couched in a phrase of maddening brevity and disturbing ambiguity. The man wasn’t quite lazy, neither was he incompetent nor unpleasant – in his employer’s words:
At first Bartleby did an extraordinary quantity of writing. As if long famishing for something to copy, he seemed to gorge himself on my documents. There was no pause for digestion. He ran a day and night line, copying by sun-light and by candle-light. I should have been quite delighted with his application, had he been cheerfully industrious. But he wrote on silently, palely, mechanically.[2]
Subsequently, upon any request for productive labour, Bartleby would repeatedly and calmly respond, ‘I would prefer not to’.
1.
Giorgio Agamben, in an essay entitled “Bartleby, or On Contingency”,[3] takes writing to be the archetype of Creation, situating his cosmic scribe in a space between the theological and the philosophical. Worlds are strung upon a thread connecting Aristotle of the De Anima to the logocentric obsessions of early Islam and Judaism, drawn through Walter Benjamin’s writing upon remembrance, and terminated in a vivid, Borgesian anecdote of Leibniz.
Leibniz’s story relates how Sextus Tarquinius, would-be king of Rome, when faced with an unsatisfactory oracle at Delphi, made for Dodona in order to lodge an appeal at Jove’s Temple, who upheld the prior judgement. The tale then moves to Athens and a dream of a pyramidal palace, of a base unseen and of infinite extent, and each of the palace’s countless rooms depict Sextus’ fate in each of the possible realisations of the world, some happy or middling, others awful. In one, Sextus heeds the god, abandons Rome, and retires as a wealthy farmer in Corinth; in another, he becomes instead king of Thrace through a fortuitous marriage. At the apex of the pyramid is the world as it is, the ‘real’, which shows what actually happened: Sextus “‘[leaves] the temple in a rage, scorning the counsel of the Gods…bringing confusion everywhere, violating the wife of his friend”’.[4]
Agamben, commenting on Leibniz, writes:
It is difficult to imagine something more pharisaic than this demiurge [the pyramid’s creator], who contemplates all uncreated possible worlds to take delight in his own single choice. For to do so, he must close his own ears to the incessant lamentation that, throughout the infinite chambers of this Baroque inferno of potentiality, arises from everything that could have been but was not, from everything that could have been otherwise but had to be sacrificed for the present world to be as it is. The best of all possible worlds projects an infinite shadow downward, which sinks lower and lower to the extreme universe—which even celestial beings cannot comprehend—in which nothing is compossible with anything else and nothing can take place.[5]
2.
Leibniz’s Palace resembles images in other theologies; for instance, Indra’s Net of the Avatamsaka Sutra in Mahayana Buddhism presents an infinite lattice of interconnected pearl-like spheres, where all worlds are reflected in all others. A rather less picturesque update, this time from contemporary philosophy, is Nick Bostrom’s Simulation Argument (SA).[6] Beyond its philosophical interest, the SA provides a fertile ground to jointly discuss existential risk, theology, and the function of the museum.
Although formally specified as a statistical argument, it can be loosely summarised as follows: if humanity doesn’t go extinct beforehand, then progress in Artificial General Intelligence (AGI), a stronger form of AI, eventually leads to a world where algorithms and computers are powerful enough that our posthuman descendants are able to simulate all or part of human history, down to the level of individual subjective experiences. Assuming these posthumans don’t prohibit human simulation, for instance for ethical reasons, there is a non-zero, albeit difficult to quantify, probability that the world we inhabit and experience currently is a simulated one.
3.
In an echo of Melville’s narrator’s puzzlement over Bartleby’s prior lives, the SA begs a question: how and why might our posthuman descendants choose to simulate their ancestors?
The theologian Eric Steinhart has proposed that posthumans may simply be entertaining themselves in some equivalent to a VR holiday to the past; they may be remembering earlier versions of their society; or, it may be some sort of artistic or cultural activity. Alternatively, they may want to test out technologies (such as increasingly powerful orders of AGI) within what, in computer programming, is termed a ‘sandbox’, essentially a controlled environment that can be shut down and doesn’t affect anything in the posthuman simulators’ ‘real’ world. Or, perhaps it is research: they want to work out what went wrong to cause some major, but presumably not extinction-level, catastrophe that is yet to befall us, in the portion of human existence that remains between the present and posthuman time.
Perhaps the posthumans engage in an act of remembrance. Remembrance is something Walter Benjamin repeatedly grappled with, most eloquently in his “Theses On the Philosophy of History,” where it is the powerful force that keeps alive the possibility, at least, that the movement of history might constitute something we might call ‘progress’. Giorgio Agamben writes of Benjamin:
Remembrance can make the incomplete (happiness) complete, and the complete (pain) incomplete. This is theology – but the experience of remembrance forbids us to conceive of history in a fundamentally atheological manner, even as we are not allowed to write history directly in theological concepts … Remembrance restores possibility to the past.[7]
Benjaminian remembrance has a pre-echo, if not an antecedent, in the Russian Cosmists.[8] Nikolai Fedorov, in a 1913 essay on the museum, writes:
Then the insensible would not prevail, it would not take the life of the sentient, then all that is sentient would be restored, and all worlds would be united in resurrected generations, and an infinite area would open for their conjunctive activity, and this alone would make internal discord unnecessary and impossible.[9]
In the contemporary context, Cosmist ideas, specifically in relation to the museum as a site of remembrance, as opposed to merely a repository for this and that aesthetic objet d’art, is developed by Boris Groys, as well as in the work of filmmaker Anton Vidokle.[10] Groys extends Fedorov by citing the museum as a post-biopolitical structure, which in some imagined future utopia “[overcomes] the boundaries between life and art [not as] a matter of merely introducing art into life but [as], rather, a radical museification of life.”[11] Groys, in another essay, places the museum in a Heideggerian context, as an antidote to technology, a site where the contemporary preserves and improves upon the old:
And, if a museum is doing its job well, [artworks] are kept in good condition, restored, and so forth. The use of the artworks is their contemplation—and contemplation leaves the artworks undamaged. Thus, turning a human being into a thing does not necessarily diminish his or her dignity. Technological self-reflection does not annul human rights but, in fact, radicalizes them by treating the human being as an artwork. However, art serves not only to conserve objects but also to improve them.[12]
Read through the lens of Bostrom’s writing, the posthumans could seek to preserve current humanity, that is, some version of what they think their own ancestors were like, through simulations, thinking of our society as a wondrous object of veneration, a cautionary tale, or in some way and sometimes a bit of both.
Lastly, on the topic of reasons for ancestor simulation, it is also possible that the posthumans have no reason that we can understand, to simulate us.[13] As suggested in sections 414-416 of Leibniz’s Theodicy, many possible renderings of the world may exist, the reasons for which exist in the world of the creator, whether posthuman or theistic, and therefore remain inaccessible to those within any specific instance. The tortured Sextus in the pyramid’s apex has no idea why he didn’t instead end up a happy farmer at Corinth.
4.
Returning now to the world, all-too-finite in time and extent, that we inhabit: remembrance and memory are inevitably bound with the objects that surround us. In fact, in many cases, it is the object that delimits time through its association with a past event.
For instance, consider the Rai or Yap Stones. Their material, limestone, isn’t found on Yap Island, and must be mined and transported from 150 nm away in Palau. These immensely heavy objects, once imported to Yap, remained, in many cases, in situ, and the community developed a consensus on who owned the stone, in whole or in part. This consensus, although verbally distributed amongst the community, was remarkably resilient:
According to local legend, two or three generations before Furness’s [an anthropologist] arrival, a large stone in transit was lost at sea during a storm. The claim to this stone continued to have value, even though the stone itself was unrecoverable.[14]
These intensely physical objects entered into and encoded specific aspects of the collective memory of their society. In addition to remembering who actually owned the Yap Stone, the community internalised the Stone’s history, which influenced their value, unlike in the case of fiat money or most financial securities.[15] Indeed, the Yap Stone’s identity and value lay not in its material, nor in its shape, nor in its location, but rather in the (collectively recollected) difficulty of acquiring the stone, combined and activated by the community’s remembrance of certain key facts (or narratives) about it.
The Yap Stones implicitly mark time by commemorating events, and, by construction, indelibly connect time with value. This notion of externalised and socialised memory, encoded in an object, is what Bernard Stiegler terms tertiary retention,[16] primary and secondary retention being respectively our experience of the world and our initial recollection of experience from memory. Slightly different perspectives come from American central banker Narayana Kocherlakota and anthropologist David Graeber, both of whom, albeit from somewhat different positions, held that money, which is a type of debt, albeit one that carries a zero interest-rate, embodies social memory.[17] Put differently, money, in both archaic (object or metal-based) and contemporary (fiat) conceptions, can be seen as a way of converting physical resources, namely labour and natural resources, into a social construct/contract in the form of debt, which allows for transmission across space and (forward) through time.
5.
The Yap Stones relate an object to a particular social narrative, that, importantly, is oral. They can be seen as a predecessor to both Bartleby, who records in writing, as well as the networked, anonymised, digital writing of the blockchain. This brings us to the last year’s hysteria around Non-Fungible Tokens, the Artworld’s foray into cryptofurore: NFTs, DAOs, Dogecoin, Ether, Polygon and, not least, Beeples.
But firstly, what is an NFT? NFTs are a species of ‘smart contract’: a self-executing, and in some cases, recursively self-modifying piece of code that communicates with other software.[18] Smart contracts reside on a blockchain, an algorithmic structure, akin to the Scrivener’s ledger or an archive’s index, that keeps track of information, specifically who owns a given digital asset, such as (in the case of most NFTs) a photo, video, sound file, or other piece of software. Blockchain can also be viewed as a ‘truth engine’: an attempt, within a well-defined, self-selected community, to establish, via a mathematical procedure, what events are agreed as being true (such as who owns an asset), and, of equal importance, to establish in what time sequence such consensually true events have happened.
Smart contracts exist natively in what philosopher Yuk Hui terms a ‘digital milieu’, the environment of software objects and logical or technical relations and usages that exist within, and enable, our computation-centred life.[19] These include various protocols and languages, such as HTML, SGML, XML, three ‘languages’ that govern the interaction of web-based objects and make smartphones, the internet-of-things, pervasive gamification, and social networks all work. He also proposes that data on the web represents a type of memory, or record of experience, experienced individually or collectively, that is external to our own memory, for which he repurposes Stiegler’s tertiary retention.
Within Hui’s formulation, the technologies underlying the blockchain can be seen as a mode of relation between increasingly autonomous, in the sense of being algorithmically-governed, entities. So, for instance, while most NFTs point to some target media – a still image or a predefined video – a few NFTs contain software that makes them display in a constantly evolving way, repeating themselves only after an arbitrarily long time, that in theory, could be hundreds of years.[20] An NFT does not have to be a pre-made image or video, it can be an active piece of software that responds to various criteria on and off the blockchain. For instance, an NFT that destroys itself as soon as it has left the public domain, that is, once it has been bought by a collector, thereby challenges the capital-based acquisitional logic underlying contemporary art.[21]
6.
The motivations and methods of capital are also embedded in the blockchain’s origin myth, as articulated in Wei Dai’s 1998 proposal that predates Satoshi Nakomoto’s Bitcoin and Vitalik Buterin’s Ethereum.[22] Theirs was a utopian ideal: that money, which had historically been connected to taxation and consequently the state’s monopoly on violence, could have a different anchor, one not dependent on forced extraction, geography, or centralised authority. Within the context of economic history, cryptocurrencies can be seen to continue the project of modern money, particularly after the 1970s: that of unmooring value from well-defined, observable, physical collateral, such as gold (Keynes’ ‘barbarous relic’ [23]) towards the fiat status quo ante-crypto where money is backed by the ‘faith and credit’ of a more-or-less powerful and trustworthy government. An important caveat: while the blockchain’s truth-finding procedure is algorithmic and transparent, and to some extent merits the extravagant claims its proselytisers are prone to make about it, it still relies – and the associated ecosystem of technologies that make it more than an intellectual curiosity – on the web, i.e. a network emphatically under state control, and that still consumes real economic and ecological resources. Put another way, cryptocurrencies are basically useless unless and until they are converted to fiat money or are used in the real economy, both of which fall under the regulatory supervision of the state.
7.
In the pre-crypto world, which for the purposes of this paragraph, we effectively still are in, truth, at least of the legal or contractual variety, would be adjudicated by courts, and enforced by the state. Thus, another utopian aspiration of the blockchain lies in the smart contract [24] itself, namely the idea that adjudication, litigation, uncertainty and cost in enforcement of legal contracts could be done away with if legal provisions could be written into unchangeable digital code.
As a tool for social organisation, smart contracts can be used to create the contemporary version of medieval guilds [25] – Decentralised Autonomous Organisations (DAOs). DAOs allow for publicly-auditable, geographically-dispersed collaboration networks amongst strangers, for instance in arts communities.[26] DAOs also offer potential to re-imagine the democratic process,[27] allowing more transparent proposal and voting [28] upon specific proposals, as an alternative to representative democracy’s prevalent mix of quasi-legal corruption, lobbying, and regulatory capture.
Unlike Hui’s milieu, which consisted of human-created digital objects inter-communicating, smart contracts and DAOs are perhaps an early iteration of semi-autonomous digital objects that interact with each other, and possibly with the broader world, recursively modifying themselves in response to their environment. At this point, these objects and networks act in accordance with human-designed algorithms and are very far from conscious objects with independent agency, but the resulting network relationships and behaviours may, at the very least, be complex. A perhaps analogous example in AI is the fact that certain machine-learning algorithms are inscrutable [29] to their creators, that is, they arrive at decisions or outputs that are not explainable in a relatively simple or heuristic way by the humans who created the algorithm, an issue that apparently arises from their complexity.
More speculatively still, consider Robin Hanson’s writing around a world populated by emulations [30] (human minds that have been uploaded to computers), the implications of which are further developed in Nick Bostrom’s Superintelligence [31]: this is a society where computer-based autonomous agents do most of the things we currently think of as productive human work. In order to do this, and assuming their society is something like ours, they would presumably need to buy and sell services, and perhaps goods (such as raw materials, electricity or replacement parts), from each other. Would they use smart contracts to make ‘business’ agreements? What would be the philosophical and practical features of their currency?
8.
The blockchain’s Platonic instantiation of time is neither like our clock-time nor like the socialised marker-time of Yap. One senses an echo of the distinction Benjamin drew between everyday calendrical time and the Messianic time of revolutionary change.[32] The temporality of the blockchain is, as Nick Land puts it, a
complex chrono-synthesis, occurring at the boundary where Bitcoin’s artificial time – proceeding only by successive envelopment (of blocks into the chain) – meets the social-chronometric time of measurable periods. “Ten minutes” means nothing on the blockchain (in itself), until connected by an interlock that docks it to a chronometer.[33]
Land’s proposal is that the definition of time has a different ontology within an algorithm. While his example applies to the blockchain, it also seems to be a logical consequence of certain versions of AGI and whole-brain emulation, where artificially-intelligent agents, operating at much faster clock-speeds than humans, see our physical world in slow motion, as it were.[34] We become doddering nonagenarians from their perspective, while they go through epochs and fight vastly destructive wars within a few human-hours.
9.
Rainer Werner Fassbinder’s World on a Wire (1973) presents a similarly destabilising vision of stacked simulated worlds, something that is permitted within the logic of the SA. There are present-day analogies: video games today simulate arbitrarily-complex worlds and richly-featured agents, where there is no reason these agents couldn’t run simulated worlds, and so on, recursively. In computer science or in the terms deployed by Bostrom,[35] we can think of virtual machines that allow a Windows machine to emulate a Mac, or a Chrome browser running on a desktop to emulate what Chrome might look like on a phone. There is also the Ethereum Virtual Machine (EVM): the basic software on which the Ethereum blockchain runs.[36] Instead of being deployed on one machine, it is spread across thousands of interconnected computers distributed worldwide. The EVM is Turing Complete, meaning it can do anything a ‘normal’ computer can do, but it is optimised for the specific task of maintaining the blockchain and adding verified blocks to it.
By analogy with Fassbinder’s film, we can imagine that there are intermediate orders of simulators between our world and what Bostrom terms the ‘basement level’ [37] that is the most fundamental level of the hypothesised computational hierarchy. Importantly, this foundational physical computer might look nothing like ours, even extending to Matrioshka Brains, the planet-sized computers that would capture all the energy of a host star to run simulations of vast worlds.[38] The planetary computer, whether in this form or the quasi-organic being in Stanislaw Lem’s Solaris (1961), is a trope of science fiction, but it is arguably inchoate in the current pervasive web, the internet-of-things, or distributed computing projects.
10.
To create a world is to write a world. Yet who is the scribe? The turn of the millennium sees networked agents-become-scriveners, consuming and generating packets of data that are now the majority of planetary network traffic.[39] Whereas in earlier decades this rhizomatic memory was split between centralised servers and localised, ephemeral memory on thousands of desktops, today the blockchain presents a territorially-unbound tabula that is continuously passed around a ring of nodes, radically upending both the task of, and constraints upon, remembrance itself. Whilst NFTs and cryptocurrencies attract excitement, the emerging nexus between the blockchain qua deterritorialised memory and a nascent AGI seems to present the prospect of a darkest dawn yet to come.
Time is both the medium within which existence, the Heraclitean flux, is suspended, as well as the numeraire against which labour, human, industrial, digital, is, since Marx, marked. Hence the ontological terror that Mark Fisher saw in Fassbinder’s film could be read as a dual horror: the film’s character Einstein is confused about his being in this level of reality or that, while the character Stiller is befuddled by the curious dilation he experiences in his, ontologically more fundamental, yet still simulated, layer. In both cases, being and time dissolve into an apparently arbitrary nihilism.
Bartleby’s case is different: by neither flatly refusing to work, nor assenting, he remains a being suspended in time, a renunciate, undergoing “the most extreme trial a creature can undergo”.[40] He, in the sense of an artist or monk, has placed himself just outside society, not quite in the desert, but in the place of exiles beyond-the-walls as the Romans would have it. Hence, while his choice can be read as a type of arrogance or privilege, there is more than a hint of the bare life about it, and indeed, he ends his days in the walled garden of the Tombs, a prison for vagrants, sleeping “with kings and counselors”,[41] gently fanned by the crimson-hued wing of the archangel Gabriel.
–
1. Herman Melville, Bartleby, The Scrivener: A Story of Wall-street, 1853 (Minneapolis: Indulgence Press, 1995).
2. Ibid.
3. Giorgio Agamben, “Bartleby, or On Contingency” in Potentialities, ed. and trans. Daniel Heller-Roazen (Stanford: Stanford University Press, 1999).
4. Ibid., 266.
5. Ibid., 265-66.
6. See Nick Bostrom, “Are You Living In A Computer Simulation?” in Philosophical Quarterly Vol. 53: No. 211 (2003): 243-255. <https://www.simulation-argument.com/simulation.html>.
7. Agamben, Potentialities, 267.
8. See Russian Cosmism ed. Boris Groys (New York: e-flux; Cambridge: MIT Press, 2018).
9. Nikolai Fedorov, “The Museum, Its Meaning and Mission,” e-flux #65 (2015). <https://www.e-flux.com/journal/65/336461/the-museum-its-meaning-and-mission/>.
10. See my review of Anton Vidokle’s work here: https://publicseminar.org/essays/silenus-cup-drained-by-ai/.
11. Boris Groys, “Genealogy of Humanity,” e-flux #88 (2018). <https://www.e-flux.com/journal/88/175803/genealogy-of-humanity/>.
12. Groys, Russian Cosmism, 8.
13. See Robin Hanson, “How To Live In A Simulation” in Journal of Evolution and Technology Vol. 7 (2001). <https://www.jetpress.org/volume7/simulation.htm>.
14. See https://jpkoning.blogspot.com/2013/01/yap-stones-and-myth-of-fiat-money.html.
15. For reference, see Cora Lee C. Gillilland, The Stone Money of Yap: A Numismatic Survey (Washington: Smithsonian Institution Press, 1975).
16. For a helpful gloss on Stiegler’s tertiary retention, see https://www.3ammagazine.com/3am/stieglers-memory-tertiary-retention-and-temporal-objects/.
17. See Narayana Kocherlakota, “Money is Memory” in Journal of Economic Theory Vol. 81: No. 2 (1998): 232-251 <https://www.minneapolisfed.org/research/staff-reports/money-is-memory> ; David Graeber, Debt: The First 5000 Years (New York: Penguin, 2012).
18. See Laura Lotti, “The Art of Tokenization: Blockchain Affordances and the Invention of Future Milieus” in Media Theory Journal Vol. 3: No. 1 (2019). <http://mediatheoryjournal.org/laura-lotti-blockchain-affordances/>.
19. For Hui’s work, visit his website: http://digitalmilieu.net/about-yh/.
20. For example, see https://www.larvalabs.com/autoglyphs.
21. See http://ukc10014.org/crypt0den.html.
22. For Wei Dai, see http://www.weidai.com/bmoney.txt; for Satoshi Nakomoto, see https://nakamotoinstitute.org/bitcoin/; for Vitalik Buterin, see https://ethereum.org/en/whitepaper/.
23. See https://www.econlib.org/ludwig-von-mises-on-the-barbarous-relic/.
24. See https://www.fon.hum.uva.nl/rob/Courses/InformationInSpeech/CDROM/Literature/LOTwinterschool2006/szabo.best.vwh.net/smart.contracts.html.
25. See https://matdryhurst.medium.com/dao-guilds-establishing-territory-e8ba64ae6f25.
26. See https://foundation.app/blog/holly-herndon-and-mat-dryhurst.
27. See Ralph C. Merkle, “DAOs, Democracy and Governance,” in Cryonics Magazine 37:4 (2016): 28-40. <http://merkle.com/papers/DAOdemocracyDraft.pdf>.
28. See http://mason.gmu.edu/~rhanson/futarchy.html.
29. See https://nautil.us/is-artificial-intelligence-permanently-inscrutable-5116/.
30. See https://www.overcomingbias.com/tag/ems.
31. Nick Bostrom, Superintelligence (Oxford: Oxford University Press, 2016).
32. See https://joncast.com/considering-walter-benjamins-homogenous-empty-time/.
33. See Section 3.45 of Nick Land’s “Crypto-Current: Bitcoin and Philosophy”: https://etscrivner.github.io/cryptocurrent/.
34. See Robin Hanson, “What Will It Be Like To Be An Emulation?” in Intelligence Unbound: The Future of Uploaded and Machine Minds eds. Russell Blackford, Damien Broderick (Wiley, 2014). <http://mason.gmu.edu/~rhanson/IntelligenceUnbound.pdf>.
35. See https://www.simulation-argument.com/brueckner.pdf.
36. See https://phemex.com/academy/ethereum-virtual-machine.
37. Bostrom, “Are You Living In A Computer Simulation?”
38. For reference, see https://www.gwern.net/docs/ai/1999-bradbury-matrioshkabrains.pdf.
39. See Benjamin Bratton, The Stack: On Software and Sovereignty (Cambridge: The MIT Press, 2015), 198-204, 278.
40. Agamben, Potentialities, 259.
41. Ibid., 271.