Patterns of a Murmuration, in Billions of Data Points
This installment of Old People Read New SFF features JY Yang’s Patterns of a Murmuration, in Billions of Data Points. Some of us—well, me, mostly—only became aware of Yang when they read Yang’s Silkpunk works for tor dot com. Yang has in fact been active since 2012, published in venues from Clarkesworld to Apex. Some of us—well, me, mostly—should have been more observant. Although Yang’s protagonist Starling is in no sense human, it turns out Starling shares something vital with its human creators. But is that common element enough to endear it to my Old Readers?
Patterns of a Murmuration, in Billions of Data Points can be read here.
This was really hard for me to get into. I don’t know what it was, but I struggled to want to read past the first couple screens.
I reread it again, post-Kavanaugh confirmation and regrettably, it’s an easier read. The machinations of the Right to achieve its targets, regardless of the collateral, the ineffective self-absorbed thrashing for the exactly right pure ideals on the Left. Now the story makes much more sense.
Perhaps it’s because I’m a woman that I read Starling as a woman. Perhaps because I’ve grown up with the trope that the only way a woman can be effective is to become cold and emotionless that I’m sighing at the presentation of another feminine voice as such. Unless of course it’s a revenge thread. Which it is.
Then again, I’m disappointed. Because after trying so hard to establish Starling as a creature (however artificial) with independent thought and choice; after giving her/it/them a strong and compelling reason to want that revenge, it abruptly ends with Tempo’s request.
Then again, in this post-Trump, post-Kavanaugh, “good luck with the coat hangers” world? I’m not sure there is a satisfying solution to the question of what do you do when the bad guys kill the people you love.
I really bounced off this one. I didn't care for the mode of telling, struggled with awkward syntax and condescending narrative tone, lost interest in what was going on before I got to the end. It did not improve on rereading. It is unusual (in my reading) in being told solely from the point of view of the AI entities.
So thinking about what I think about it...
Looking past the style to the story, did I understand what was going on correctly? I think so. Self-aware AI entities in a high-tech, politically divided world, with access to the human world's big data, taking action to protect their interests, defying human expectations and control. The AI entities have understood the pointlessness of human ideologies and disavow political allegiance, they say, but have been assigned to action on behalf of the Left, against the Right - I do not find this credible. Do they understand themselves? They have somehow (how?) internalised their originators as parents, and their duty when the parents are threatened as revenge. So how does that relationship work? Why are they solely responsible for their mother as she is solely responsible for them? That's not a relationship I understand, in either direction, and I can't see where it comes from. (Where it would come from is an interesting story - I think Justina Robson told it once, in Silver Screen). There is a dissonance here, for me, that destroys whatever enjoyment I did have in the ideas being worked with. I just don't believe in these AIs, they are not credible persons to me. I just don't care. Not even the author killing their mother made me care.
There is an interesting world here - has JY Yang set other stories in it? - reading them first might have helped. In another mood or another life I might have reacted differently, might have found this interesting and engaging and even moving - it has that potential. But not this week in this mood, in this life.
The use of the first-person plural voice is unusual and raises some hackles. Starling is apparently a plural-core AI? Near the end, the description of what Starling “registers” in the time it takes Temp to walk down the path to the minimack, suggests massively plural cores with near-omniscience.
In summary, not much happens: an agent of the Left meets an agent of the Right, there is some banter and some conflict, and in the end, nothing is changed.
The story takes an interesting political tack, starting with the apparent assumption “Left = Good, Right = Bad,” then subverting that with Starling’s determination that “we see no difference” – from a pure data-analytic point of view – “between Left and Right.” The culminating view is Orwellian, power existing for its own sake; Nineteen Eighty-Four with a two- rather than a one-party system.
In short, I liked it, but at a brain-level, not a gut-level. It was a bit cold and, while there is the implication that Starling has emotions, they are not given to us very, well, emotionally.
Good, but inadequate to the task of actually imagining what an AI will be.
The first third requires that lovely SF provisional reading --- waiting for the details to coalesce, the rules and contours of the world to be revealed. One holds the story, tentatively, in the mind's eye, waiting for focus.
It is well done. The last part -- conflict, action, resolution – is also technically well done as a piece of writing; but significantly less satisfactory conceptually.
The anthropomorphic AI is unconvincing, beginning with the central conceit that it has a cohesive sense of self, an internal 'I'. Strutting and bragging its superiority --- in an internal monologue, oh dear ---- it manifests out of some US genre comic film --- displaying human like emotions, and not very smart human like analysis, and none of the particularity, none of the alien behaviour one can expect or imagine the future will grant us.
It is perhaps best to think of this piece, published in 2014, as a museum exhibit -- before deep learning hit our phones and Go boards, before the metaphors and contours of emerging AI became quite so frighteningly articulated.
(I should write about the language and the way Yang signals distance and tries to evoke otherness with their prose; alas work calls, and I will not).
Yang writes engagingly, but their equating of left and right is political masked as apolitical; and the resolution, which is presented, quietly, as a triumph of emotional bonding and personal loyalty over a colder revenge, is strangely regressive.
A sturdy well executed story, even if it is a deep failure in other regard
The story “Patterns of a Murmuration, in Billions of Data Points” has silken-smooth prose and a wonderful rhythm of words. I was gripped by the need to know what came next. When I found out, I wanted to argue with every word.
The narrator, an artificial intelligence or a collection thereof, determined that a horrific mass casualty event was a political plot. The narrator is proud to have figured this out; mere mortals wait until a conspirator sells the story to the tabloids.
I think what ruined this story for me was the term “cloud.” In this decade “cloud” is a piece of marketing puffery applied to internet services so that someone can resell them at 400% markup. Artificial intelligence is a self-driving car that runs over pedestrians. If we ever get thinking machines, they’ll still run us over out of a literal-minded refusal to sort through our conflicting instructions.
The conclusion that there is no difference between Left and Right is unfortunate after the last couple of weeks in the real world. Maybe nihilism is appropriate for a machine intelligence that hasn’t got the vote.
A story about the childhood of an AI. Caught up in a fight between the “Right” and the “Left”, one of its mothers has been murdered during a staged “accident” intended to take down a politician. The AI has residual loyalty only to its remaining mother, but what is the best way to respond to their loss?
The story succeeded in making me feel sympathy with the new entity and suitably disgusted/bored with the old paradigms of right, left and reflexive violence that got everybody to where they were at the beginning of the story. I give it a thumbs-up in general but my internal quibble mode didn’t shut up all the way through it. Does something that sees through all its peripherals and fights off clouds of hacking nanites really need its mother to type instructions in ALL CAPS into a “chat interface”? Why does the story take as a given that the AI’s mother’s death is more important than its regular concerns? But maybe I should just shut up and take it for what it is, a well-written story with a fresh approach to an old trope.
We are investigating a murderous blast at a political rally. We identify the perpetrator readily enough. But the question who is “we”? has already become more interesting.
Jy Yang’s story uses many quick, light details to sketch “from the inside” a community of hardware-embodied and virtual AIs against a background of simmering, low-grade civil war among the humans. By the time we get any explicit information about how this near future came to be, we care with surprising intensity about what the AIs can learn from their creators, and vice versa.
This is deft, economical world-building, storytelling, and speculation: highly recommended.
I'm not too fond of this story-- it seemed tiresome that the sides were left and right, and this was only partially redeemed by it being revealed that it didn't matter who won.
I didn't care much about the viewpoint character.
The other stories in the Old People Read New SF series had some distinction, but this just seemed pretty average.
More death. This story selection seems morbid. Or perhaps that's just me.
This one starts with a death, or rather several: a stadium disaster, that seems to the AI protagonist to be more than a accident. But only one death matters to them, that of one of their developers, trainers, parents. They've learnt human emotions well, though they claim no interest in mourning; revenge is what the transhuman sapience requires.
Quite a cyberpunky story, I felt, though with some odd anachronisms: the political factions are labelled 'Left' and 'Right', though there doesn't seem to be much resemblance to the Left and Right I'm familiar with; and a comparison is made at one point to an AK-47, surely an antique weapon at whatever time and place this story is set; might as well reference the cyclic rate of a Maxim gun.
Overall, though, I liked it. Perhaps I'm nostalgic for the cyberpunk fiction of my youth, as a distraction from this absurdist cyberpunk dystopia we're in.
This is a stylishly-written story about the birth of Skynet. The style mostly carries the reader safely over plotholes and glosses over inconsistencies: the melodramatic insistence of the assassin Rée on returning to the scene of the crime just so he can meet Starling, even though he has no idea who he is speaking to until she introduces herself; Rée’s dismissal of Starling as a “pumped-up pet AI” even though she is clearly immensely powerful and can remotely operate through artificial bodies and machinery (it’s not clear if there are any other AIs, since Starling is described as the result of a decision to give “the cloud” intelligence); why, if no one dares even question the Right’s narrative surrounding the obvious assassination of the leading-in-the-polls opposition presidential candidate, there was going to be an election at all.
This story is either shockingly naive or utterly cynical about politics, which has been reduced to a purely tribal Left and Right. (The supposed Center seems to be both powerless and meaningless.) In the end, the assassination of Hartman is of no consequence, as there is no difference between Left and Right and it doesn’t matter who is elected. In the end, the stay of execution won by Tempo will prove only temporary, and the contempt that Starling exhibits for humanity won’t be held back. None of it matters.
This is a story of a youth, Starling, comforting and aiding the survivor, Tempo, of the couple that has raised it. They were engaged in a bloody struggle against the “Right”; one, Avalanche, of the couple was taken and destroyed. Starling scouts the enemy scene.
The sci-fi element of the story is that Starling comprises a swarm of intelligent, interacting parts. The story paints the operation of this collection well. The pronouns are plural but the opinions and the actions are singular.
Starling encounters another, Wayne Rée, identified as the enemy and also an assembly of interacting, aggressive elements that attack Starling and prevail for the moment, destroying the human form, but not the elements.
Starling retreats, regroups, fights off doubts from its allies, and tries again, this time commandeering an assault vehicle to better impose its will. This time it prevails, destroys Wayne Rée, and rescues Tempo, who has rather mindlessly ridden off to a futile attempt of her own.
The story is a simple, good-guy wins short story. Its virtue lies in its representation of the bottom-line operation of the nano-bot swarms.
I’m somewhat puzzled as to the point of this story. There’s various messages that I get out of this one, but I’m not sure which of them I’m supposed to get, or whether I’m missing the one that was actually intended.
Like the others in this sequence of stories, Patterns of a Murmuration is well-written; the prose flows well and I have no trouble following it. At the same time, it doesn’t have… I don’t know, the *energy* I associate with some of the others, particularly “The City Born Great”. Nor does it have the connection with the characters that I felt in “Tongtong’s Summer” or “Seasons of Glass and Iron”. I can sort of sympathize with Starling, but with sideways messages like “no difference between the Left and the Right” and the internal thoughts of Starling often echoing “vast intellects, cool and unsympathetic” despite their obviously emotional focus on vengeance against the murderer of their mother I am left to sympathize at a distance, with a considerable suspicion that what I’m seeing is just a lead-up to what will be another AI-spawned war against humanity. Tempo may have stopped Starling from committing their murder today; I am left unconvinced that she has done so permanently, or that Starling has come to understand *why* Tempo stopped her, and why that matters. If Starling accepts no controls on their behavior, while viewing themselves as superior, this implies a very bad ending for everyone in the long run. Especially since the words of Wayne on the other side indicate that AIs like Starling are not unknown at all; if Starling is one kept under reasonable control by humans they sympathize with, well, how many AIs out there are who might go OUT of control any day now?
Overall, I think I enjoyed this one the least of the series so far. All the others had one character or another I could really sympathize with, and none of them made me feel so distanced.
What happens when a child loses a parent? Seeks revenge? Disobeys an order? What happens when that child isn’t flesh, though its parents are, and isn’t machine, though its countless bodies are, but instead a convolution of near infinite amounts of data, simulations, and heuristics? What happens when that child, living so much faster than bone and blood, becomes an adult in mere hours?
The language is brisk, fast, concise, and while suggestive of machines, still has a mellifluousness harkening to the basic human-led neurolinguistic programming our first-person protagonist was undoubtedly given at its electronic natality by its “mothers.” But how quickly can such a child take the programming it’s given, the experiences it leads, and the data that it has access to, and mature farther than any child has yet before?
And put this invented, and yet independent, person into a terrible situation. A panopticon world in a political maelstrom, Left vs. Right, both sides teetering on the edge of victory and simultaneously mutual annihilation. An accident kills hundreds, including one of the mothers… or was it sabotage? Our child investigates and uncovers the truth by stages, and along the way the question the mothers never thought to ask, much less answer: what are humans to a machine?
What choices can it make? The limbs and tools at its disposal, as a networked intelligence, are everything that is online, from cameras to lights to the weapons of modern warfare. But given this power, and this situation, and the ethics it was taught, what morality will it develop?
Is revenge digital? Is care logical? Is love rational?
I struggled writing this review. The story is well written - the central character's otherness is well-defined, and the exploration of AI motivation is satisfactory. But it doesn't move me. I get it - the factions pursuing a violent political conflict are unimportant, the possible purposes for creating an AI are immaterial to the entity in question, and the AI's motivations don't have to match our own to produce a similar, human-seeming result.
I enjoyed the style, which I have experienced before; a first person reaction to tragedy, from a collective person, who is clearly not human. But this viewpoint should tell me something that a different viewpoint could not, and I did not find the motivations and priorities compelling. I like alien viewpoints. I like stories about AIs. I love explorations of morality and motivation. This story has all of those elements, but it leaves me cold. Perhaps in the end the AI was too human.
A society sharply divided between Right and Left. Violence. Big Data. Where have I heard all this before? Yet JD Yang's story quickly leads me away from Trump-administration parallels into its own weird reality, in which human form can be assumed and put aside like a costume, using whatever interface with the world can gather the most information. We start with a tragedy -- the death of someone named Mother, and also Avalanche -- and quickly become immersed in a cat-and-mouse game of political assassination and artificial intelligence so highly evolved that it can override human commands. It's a bleak and alarming world, and perhaps the current state of the world has made me resistant to the charms of dystopia, but I found the ultimate message facile.