It is Sunday 18 October 2026. The article below was written May 21 2014. Oh if I had only known what nightmare year I was headed for. End 2014-2015 was the most daunting year in my life. Then, if only I had realized that in 2026 I’d be having an actual Artificial Intelligence rewriting this article. Absurd! The rewrite comes first, the original crappy article comes right after. I have reposted the this one today.
Eusingularization
On the plurality, distribution and political consequences of the technological Singularity
Originally written 21 May 2014. Revised and expanded.
What is Eusingularization?
The term implies two things. First, the prefix Eu-, meaning good, beneficial, or desirable. That simple prefix already introduces a distinction conspicuously absent from most discussions of the technological Singularity. There is no reason to assume that the Singularity, should anything resembling it occur, must be universally beneficial, universally disastrous, or even particularly comprehensible to the human beings living through it. There isn’t necessarily the Singularity. There is a range of possible transformations, each with radically different consequences.
Singularity is a bit like casting a 17th level D&D Wizard spell over Civilization, and specifying conditions. “I cast Singularity” – and then you spend 10 minutes carefully explaining to the Djinn the Dungeon Master has appear what kind of Singularity. The Djinn looks suspiciously like Anders Sandberg with an Ipad taking notes, asking twenty questions.
(…)
The second implication concerns my preference for Singularization over Singularity. The conventional term suggests an event, a historical boundary that humanity approaches and eventually crosses, a bit like crossing an event horizon. Before the Singularity we have one kind of civilization; afterward we have something fundamentally different. I regard this as unnecessarily restrictive. Singularization is better understood as an extended, uneven and potentially discontinuous transitional process, with different parts of humanity entering different phases at different times – a bit like spaghettification (the violence OF the Singularity ending up killing you), or entering the accretion disk or “eddingtonization” (the violations before the Singularity pushing people violently away FROM the Singularity), etc. etc.
Indeed, Singularization may already have begun all around us and have become mostly irreversible. The idea this is irreversble would anger a lot of people – “we havent voted for this!” in much the same manner as “we haven’t voted for eight lane highways” or “we haven’t voted for ozone depletion or nuclear weapons!”. No that stuff kinda happened. And now we have to deal with it.
We may already inhabit a historical period in which technological development increasingly outruns the capacity of individuals, institutions and entire societies to understand or anticipate its consequences. The distinctive feature would not necessarily be the arrival of one spectacular invention but a sustained contraction of our ability to predict the future. Developments that previously required generations may unfold within decades, then years, while their implications become increasingly difficult to comprehend.
If there can be a good Singularization, an Eusingularization, then there must also be other possibilities. I propose DySingularization for transformations producing predominantly harmful experiential conditions for human beings, ThielIngularization, where predatory and fascist elements want to proactively eliminate large sections of humanity and benefit the Singular pleasures for themselves and a small selection of those they seem worthy; Xenosingularization for transformations that become profoundly alien or incomprehensible, and Teratolarity for particularly horrifying outcomes involving overwhelmingly powerful and potentially hostile artificial intelligences. For completeness, we might even imagine the jocular Santalarity, in which technology dispenses apparently limitless happiness and presents to everyone forever.
Yes, I am aware that I have somewhat bastardized Greek and Latin. I have more pressing concerns.
The important distinction is that these categories are neither mutually exclusive nor necessarily absolute. A Singularization can simultaneously be beneficial, destructive and profoundly strange. Indeed, this may be more likely than any internally consistent utopia or dystopia. We should therefore stop describing possible futures exclusively as good or bad scenarios and begin treating them as combinations of independently varying conditions.
For convenience, imagine three scales, each running from zero to ten. One represents the beneficial consequences of technological transformation, another its destructive consequences, and the third its degree of unfamiliarity or alienation. Clearly these are not scientifically established measurements. They are conceptual instruments for discussing states that otherwise become difficult to compare. Nor are they percentages that must add up to a hundred. A society could simultaneously experience Good Level 7, Bad Level 8 and Weird Level 9. There is no mathematical contradiction because the three values describe different properties of the same experience.
Now introduce geography and demography. Imagine a region of West Africa undergoing exceptionally rapid technological development. New medical technologies, inexpensive energy, automated agriculture and radically improved communications might produce beneficial changes corresponding, very approximately, to Good Level 6. Simultaneously, labor displacement, economic dependency, predatory technological monopolies and institutional instability might generate Bad Level 5. Rapid alterations to established customs, relationships, religious practices and social expectations might produce Weird Level 4.
Yet these numbers would describe an abstraction that conceals enormous differences. Affluent, technologically connected inhabitants might experience overwhelmingly beneficial changes, while poorer populations experience Bad Level 8, potentially involving the effective destruction of their livelihoods or their ability to function within the emerging economic order. A new technological infrastructure might make a region collectively wealthier while rendering substantial portions of its population economically redundant. Both statements could be true at the same time.
Even the same individual might experience benefits and catastrophes simultaneously. Someone might gain access to astonishing medical care, unprecedented information and material abundance while losing employment, social status, privacy, familiar relationships and any meaningful sense of participation in society. Such a person is neither living in a utopia nor straightforwardly inhabiting a dystopia. They are experiencing a contradictory transformation whose consequences cannot be reduced to a single verdict.
There is also an important distinction between technological weirdness and technological harm. A future can be profoundly alien without being particularly cruel – think living as an immortal being in a never ending rapid succession of fasted paced games or video clips or hentai porn directed by another director every few minutes. Conversely, an apparently familiar society might become exceedingly oppressive through technologies that are almost inconceivable.
An automated system deciding access to employment, housing, insurance or public services may produce relatively little immediate cultural strangeness while fundamentally altering the distribution of power. Weirdness concerns the distance between familiar human expectations and emerging realities. Badness concerns the suffering, deprivation, coercion or destruction these realities produce. Neither reliably predicts the other.
I have tried explaining this through Gantt charts, shock levels, overload levels and even generalized measurements of psychological unease or dysphoria. The central idea is to visualize multiple, overlapping transitions unfolding at different speeds across different populations. One group might experience several years of moderately beneficial change followed by a sudden economic collapse. Another might suffer considerable initial disruption before acquiring institutions that allow it to benefit enormously from the same technologies. A third might remain materially comfortable while becoming increasingly alienated from the civilization around it. Vinge, Stross, Gibson, Stephenson, Egan wrote all about this, I need but look sideways at my unfinished book case.
Such a chart would not forecast history with numerical precision. Its purpose would be to expose the inadequacy of treating humanity as one homogeneous population traveling toward one universal outcome. Each region, demographic group and historical period would have its own changing profile. The scales would need to account not only for the magnitude of technological change but also for its speed, duration, reversibility and the capacity of affected populations to adapt. An intense disruption lasting three months is not equivalent to a permanent condition of social exclusion.
Unfortunately, even describing such possibilities to people in a cafe here in Amsterdam can provoke remarkable hostility. I have repeatedly encountered people who interpret attempts to analyze potentially terrible futures as expressions of sympathy for those futures. Describe a society in which algorithmic processes determine economic relevance, and someone assumes that you are advocating a hierarchy of human worth. Discuss the possible emergence of populations rendered economically unnecessary by automation, and the conversation can abruptly become an accusation that you wish such populations to disappear. Attempts to describe mechanisms of systemic exclusion are sometimes interpreted as proposals to construct those mechanisms. Just ask Yuval Harrari. He writes a book about historical and is summarily accused if being in bed with Klaus Antichrist.
This is an extremely dangerous intellectual confusion. Analysis is not advocacy. Prediction is not prescription. Identifying a possible mechanism of oppression is not an endorsement of oppression. Quite the contrary: understanding how a disastrous future might emerge is among the necessary conditions for preventing it. We cannot meaningfully oppose a development that we refuse to describe because its implications make us uncomfortable.
Some of this emotional resistance is understandable. Technological discussions frequently conceal political interests, and history supplies ample reasons to distrust those who describe human beings as expendable components of supposedly inevitable progress. But the appropriate response to a disturbing model is to interrogate its assumptions, evidence and implications, not automatically to attribute its imagined consequences to the desires of its author. A sufficiently sophisticated civilization must be able to discuss horrifying possibilities without confusing discussion with endorsement.
This brings us to the political implications of Eusingularization. If technological transformation is neither singular nor uniformly distributed, then the question is not simply whether humanity will experience a good or bad future. We must ask who benefits, who suffers, when these consequences occur, how they interact, and which institutions possess the power to influence their distribution. An Eusingularization cannot meaningfully be declared successful merely because aggregate prosperity increases while entire populations become helpless, dispossessed or irrelevant.
Someone with extraordinary power whispers to the US president in a careless moment of excess ketamine use, some service is throttled, thousands of people die. Poof, warcrime.
Nor should we assume that desirable outcomes emerge automatically from greater technological capability. Increased productive power does not necessarily produce increased freedom. Greater intelligence does not necessarily produce greater compassion. Extraordinary abundance does not guarantee equitable access, and the ability to solve ancient material problems does not prevent the creation of entirely new forms of dependency and coercion.
I therefore strongly advocate active personal and political involvement in the process of Singularization. If anything resembling such a transformation is occurring, then the choices made during its development may determine the circumstances under which generations of human beings, and potentially other conscious beings, must live. The more powerful the technologies involved, the more consequential seemingly mundane questions of ownership, governance, accessibility and human rights may become.
An Eusingularization is not a happy ending promised by technology. It is an ongoing political, ethical and civilizational objective: the deliberate attempt to maximize desirable outcomes, minimize destructive ones, and preserve meaningful agency amid changes that may become increasingly difficult to comprehend.
We may not be able to prevent the future from becoming strange. We may not even be able to comprehend all the transformations occurring around us.
But we can refuse to confuse technological inevitability with moral inevitability. We can insist that the consequences of progress remain legitimate subjects of political struggle, and that the people most exposed to its dangers are not simply written off as unfortunate collateral damage.
The future does not arrive everywhere at once. It does not arrive equally for everyone. And it does not ask permission before becoming incomprehensible.
That is precisely why we should concern ourselves with making it good for literally every stakeholder. Leave people out, or even imply you might not have anyone onboard, and you will immediately risk make people very angry. You might end up under something worse than a Guillotine as a world leader one day.
With absolute power and ultimate existential stakes comes absolute responsibility
What is Eusingularization?
This term implies two things, first – the prefix Eu- for ‘good’, or “desirable”. That simple prefix conceptually implies that there is a range and diversity of possible “Singularities”. There isn’t “the” Singularity. Anyone who keeps talking about “the” Singularity simply doesn’t get it.
The second implication of my use of the term Eusingularization is the more diffuse term of Singularization versus Singularity. In the typical phrase of the concept of Singularity, it’s an event. To oppose this line of thinking I advocate the more fluid idea of a transitional process of Singularization.
“Singularization” may have already started in the past. We may in our current society be in a transitional Singularizing stage of human history, where increasingly the future becomes an opague or unpredictable state in the increasing near future. In other words, the further in to the future the more unknowable the near future becomes.
So if a Eusingularity is a good transition towards what comes after, there should also be things like a Dysingularity (A Singularity that introduces mostly bad experiential states for the current human species), a Xenosingularity (where humanity loses all relevance and progress becomes incomprehensible, i.e. where an appreciable part of humanity survives, but those that do no longer have any systemic relevance), a Teratolarity (where terrifying artillects call the shots and surviving humanity is left at the mercy of horrible powers). For completenes sake I mention the jocular Santalarity (A singularity where humans get presents and happiness for all eternity).
Yes I am aware I somewhat bastardized greek and latin here.
In my introduction of the phrase EuSingularization I advocate personal and political involvement of people towards the Singularity stage in human history. If this is a true concept, people should be compelled towards personal interest and involvement. The idea that we may be in or very close to this transitional stage in history means we may have moral culpability or responsibility this event ends well for all involved.
References:
http://www.etymonline.com/index.php?allowed_in_frame=0&search=singul&searchmode=none