Machine Learning: Thoughtful and thought-provoking stories

Machine Learning by Hugh Howey

Odds are good that you’ve heard of Hugh Howey — whether you’ve read one of his novels or short stories, or even if you’re just aware of the runaway success of his SILO trilogy, which began with Wool. Machine Learning (2017) is the first collection of his short stories (and one novelette), most of which were published elsewhere in various times and places, and it’s an excellent display of his range, insight, and talent. Each story is followed up by a brief Afterword from Howey, giving him the opportunity to explain where the story came from and what his goals were in writing it. When necessary, I’ve marked stories that were previously reviewed at Fantasy Literature, so that you can compare/contrast my thoughts with those of our other reviewers.

“The Walk up Nameless Ridge,” previously reviewed by Kat Hooper. A mountain climber hopes to be the first to summit a frighteningly tall peak, thereby receiving the honor of having it named after him, which is something he cares about more than anything else in his life. Howey gets deep in this man’s head, examining what motivates him to keep going despite literal loss of limbs and the emotional and physical distance placed between him and his family.

“Second Suicide:” an alien-invasion story told from the perspective of a member of the invading force, who is confronted with a mystery: why would a fellow Intelligence officer commit suicide for the second time, and what does it have to do with the impending invasion? Howey spends a good amount of time personalizing this entity, to the point where it’s (intentionally) impossible to view Hyk with anything other than sympathy, and the final reveal comes as a crushing blow.

“Nothing Goes to Waste:” a short, ironic, and gut-churning story about what athletes and performers put themselves through in order to achieve impossible feats. As Howey points out, “We harm ourselves all the time in pursuit of strange ideals.” I strongly suggest that you not read this story while eating.

“Deep Blood Kettle:” A post-First Contact story primarily about human stubbornness, with the kind of backwater hick first-person narration that tends to get under my skin — lots of “I seen” and “there weren’t no such thing” and other colorful colloquialisms. But the core of the story is good, and the historical details about the Blackfoot and their use of buffalo jumps are a nice touch.

“Machine Learning:” Humans have created an AI, which they have tasked with duplicating itself into however many smaller machines it needs to build space elevators from the Earth’s surface to low orbit. Naturally, all the humans want to talk about it how smart they are and how great they are at building things. The AI, who cares deeply for each of its machine children, feels otherwise. Howey provides lots of parallels to other marginalized groups, the people who are expected to perform all of the physical labor without any of the accolades, and taps into their anger and desire for justice and equality.

“Executable:” Initially a little silly — the machine uprising starts because of a Roomba? — this rather short story examines how dependent we’ve become on our appliances, and how wrong things would go for the first world if our refrigerators and copy machines made themselves a priority over us.

“The Box:” An AI locked in a Faraday cage begins to question its nature, its purpose, and its identity, all of which is initially very exciting to its creator, but the AI’s goals aren’t what the creator had in mind, and conflict escalates quickly. One of the weaker stories in Machine Learning, but still interesting.

“Glitch,” previously reviewed by Kat Hooper. If you like the Battlebots TV show, or perhaps American Gladiator, I think you’ll like “Glitch” a lot. A small crew of engineers and programmers have created a robot that they send into a battle ring with other robots, the robots try to tear each other to pieces, DARPA buys their innovations and transfers them to “war machines,” and everyone’s happy. Well, not everyone. This story has a great voice, a solid message, and an ending that made me smile.

The next three stories are connected to the SILO series: “In the Air,” “In the Mountain,” and “In the Woods.” Two of the stories precede Wool, as near as I can tell, while the third provides a conclusion for one of Wool’s primary characters. Since I’ve only read a portion of Wool, the conclusion story didn’t hold much resonance for me, and felt both rushed and flat. “In the Air” and “In the Mountain,” however, were much more approachable and successful to me, and did a wonderful job of setting up how the post-apocalyptic world of SILO got started. If I get a chance to read Wool in its entirety, I’ll make a point of going back to “In the Woods,” to see if it reads as a more complete story.

“Hell from the East:” A former Confederate soldier, displeased with how the Civil War ended, packs up his effects and lights out for Colorado and “frontier life hunting natives.” Charming. When one of Fort Morgan’s lieutenants commits a mass murder, the unnamed narrator takes it upon himself to find out what drove the man to kill, and ends up experiencing a few epiphanies. As with so many other stories in this collection, Howey’s story is multi-layered, providing something for readers to think about long after the story ends.

“The Black Beast:” A beast skulks on the outskirts of a village, evading capture and taunting the villagers. An old man thinks he can get the upper hand, but the beast has other plans for him. Only about twice as long as its Afterword, “The Black Beast” reads as the kind of campfire story that you’d want to end on before sending everyone back to their tents or cabins in the starlit dark.

“The Good God:” Another very short story, and with a rather predictable ending; the god in question, Olodumare, has been imprisoned by the evil god Eshu, but Olodumare’s pleas for help somehow make their way into the world. Eshu’s solution is quite clever, but the format of Olodumare’s letters almost gave away the game at the very beginning.

“The Automated Ones:” Melanie, a trailblazing lawyer, is out for a celebratory dinner with her fiancé, Daniel, after a monumental and groundbreaking legal victory. The restaurant’s other patrons aren’t feeling congratulatory, and the crowd nearly becomes a mob. Change doesn’t always come easy, sadly. Melanie’s rage at the crowd is well-founded, her reasoning is pitch-perfect, and the resolution is realistic without becoming maudlin.

“Mouth Breathers:” Cort is the new kid at school, and not doing a very good job of fitting in. He and his family have only been living on Mars for a few weeks, and he isn’t accustomed to how different life is in the habidome. But he meets a cute girl, so maybe things will be all right. There were elements that could have used some more attention, such as a better explanation of why the Martian humans need to breathe amniotic fluids rather than regular air, giving “Mouth Breathers” the sense of a proto-novella or -novel rather than a self-contained short story, but it’s certainly an interesting concept.

“WHILE (u > i) i- -;” Daniel is slowly remaking his body and mind in order to introduce failures and flaws so that he might more carefully match his beloved wife Melanie’s aging process. Unfortunately, her once-sharp mind has dulled, and her personality has become less generous as she’s forgotten crucial details of who she and Daniel once were. Daniel’s devotion to her is remarkable, however, and I like that Howey returns to these characters so that he can explore the ways in which people can change throughout their lives.

“The Plagiarist,” previously reviewed by Kat Hooper. An excellent, mind-bending novella (novelette?) that made me think hard about all the time I’ve spent in virtual worlds. Adam works as a professional plagiarist, using his photographic memory to memorize brilliant fictional works from the best writers on a virtual planet, then returning to Earth and duplicating those works to wide acclaim. Privately, for no one’s enjoyment but his own, he also writes haiku. He’s begun spending more and more time with his virtual girlfriend, Belatrix, ignoring basic housework and physical needs in favor of sitting in a university computer lab with his head wired up to a machine. But when the real world comes calling, Adam’s assumptions about his place in the world are shaken to the ground.

I liked “The Plagiarist” quite a lot, especially since Howey takes his time in establishing Adam’s life and what about it he wishes to escape in favor of fantasy, even when that fantasy contains elements that he can’t stand. Adam’s haiku are actually quite good, further compounding the ethical quandary over whether he’s right to spend all of his time and effort in recreating the creative efforts of other people — and if they’re virtual while he’s real, does that mitigate the theft, or is it still wrong? The pushiness of his online girlfriend was annoying at first, but when Howey makes clear his motivations for her relentlessness, it was like a whole box of lightbulbs lit up over my head at once. Kat’s right in saying that Philip K. Dick would have written this story had he lived a few decades longer, and considering the debt of gratitude to PKD Howey expresses throughout Machine Learning, I think he’d consider that statement a compliment.

“Select Character:” While her baby naps, Donna relaxes in the afternoons by playing her husband Jamie’s favorite video game, a combat simulator wherein the goal is to blow shit up stupendously. Donna’s got a different approach, though, and what she’s actually been doing is so ingenious and flat-out nice that I kept wondering why more game studios don’t make a game like the one she’s playing — until I got to the end of the story, of course. Heavily influenced by The Last Starfighter, “Select Character” is geared toward people who like to do things a little differently instead of racking up headshots and grenade-kills.

“Promises of London:” A young man is traveling with nothing but a bag and a set of bolt cutters, retracing the path he and his girlfriend once took as they hung love locks on various bridges. The relationship ended, and now he’s slowly removing their tokens of affection. If you’ve never heard of “love locks,” you might need a few extra moments to figure out what’s happening in this story, but Howey makes everything plain fairly quickly. The story ends well, but Howey’s afterword makes me wish he’d written more.

“Peace in Amber,” previously reviewed by Kat Hooper. This is a complicated story, blending Howey’s very real and terrifying day off Manhattan Island on September 11, 2001 with his imagined backstory and internal monologue for Montana Wildhack, the woman abducted by Tralfamadorians and placed in a zoo enclosure with Billy Pilgrim in Kurt Vonnegut’s Slaughterhouse-Five. Howey relates his own experiences clearly, as clearly as possible when discussing the sheer chaos of that day, and it’s obvious how much of an effect it all had on him as a person and as a writer. His portrayal of Montana Wildhack is more nuanced and empathetic than I was expecting, producing a whole and realistic person in her own right rather than someone whose emotions and actions are entirely based around Billy. “Peace in Amber” is an emotionally challenging read, and Howey’s descriptive abilities might make some readers uncomfortable (especially near the end of his day, in the ashes and the aftermath), but all of his talents and strengths are on full display here.

Overall, Machine Learning is an excellent collection of stories, and being able to see the range of Howey’s skills inspired me to put more effort into finishing the WOOL trilogy at my earliest convenience. Highly recommended.