Modal Title
Machine Learning / Tech Life

Artificial Intelligence: The Work of AI Satirist Eve Armstrong

Artificial intelligence may be able to change the world, but it won't help you land a prom date with Barry Cottonfield, an AI researcher finds.
Apr 1st, 2023 6:00am by
Featued image for: Artificial Intelligence: The Work of AI Satirist Eve Armstrong

For nearly eight years I’ve written tech features — over 400 of them — for The New Stack. But one of my all-time favorites was the tale of what I’d described as “a new research paper with an intriguing title” about artificial intelligence that “caught the attention of the scientific community” back in 2017.

Right around April 1

But of course, it was more than that. And less… The paper’s title was “A Neural Networks Approach to Predicting How Things Might Have Turned Out Had I Mustered the Nerve to Ask Barry Cottonfield to the Junior Prom Back in 1997.”

Eve's prospective date for the 1997 prom

Fig 1.

The author promised she’d applied the resources of a feed-forward artificial neural network to all of the available datasets. (Although most of them were ultimately derived from her five-volume high school diary…)

The paper’s author was Eve Armstrong, then a real-life postdoctoral researcher at the University of California San Diego’s BioCircuits Institute who’d already published over 20 (real) scientific papers, including “An Optimization-Based Approach to Neutrino Flavor Evolution.” It was a thought-provoking paper that last year she expanded into a real-life astrophysics presentation at the University of British Columbia.

Although the second half of that talk was still titled “A Neural Networks Approach to Predicting How Things Might Have Turned Out Had I Mustered the Nerve to Ask Barry Cottonfield to the Junior Prom Back in 1997.” (“That’s it for the neutrinos,” she begins by telling her audience….)

Those neutrino explorations had relied on a technique known as machine learning inference, and Armstrong tells her stunned undergraduate audience, “I’d like to demonstrate to you how versatile this inference technique can be…”

And then at the 41-minute mark, she veers off into alternate applications of the methodology in the generation of “an artificially-intelligent means to escape discreetly from the departmental holiday party.”

Eve Armstrong talk - Optimization predicts neutrino flavor evolution

In writing about Armstong, I’d eventually decided to declare her an “AI satirist.” At the end of her talk, she thanked her professional colleagues, her collaborators, and Barry Cottonfield, who’s “been inspirational to me throughout my years in science.”

The most remarkable part isn’t that she then fielded 18 minutes of questions from her undergraduate audience. It’s that nearly all the questions somehow sidestepped her interest in Barry Cottonfield altogether, delving instead into the details of neutrino interactions with atmospheric matter.

Although one undergraduate did ask her what ultimately was the best way to escape that Christmas party… (“Keep an eye on which exits seem to be more sparsely populated,” Armstrong advises…)

“The whole Barry Cottonfield thing — It’s almost like a paradigm for how you do machine learning,” the class’s instructor says, giving his students a tongue-in-cheek hint. “It’s almost like if he didn’t exist, you might’ve just made the whole thing up in order to illustrate machine learning.”

And Eve tells her audience of undergraduate students about its utility as a teaching tool. “I had my non-science friends actually read it, and now they know a little bit about machine learning.”

Whereas, “They would never read my neutrinos papers.”

But eventually, you realize Eve Armstrong is a real heavy-hitting scientist whose research interests included the astrophysics of neutrinos and biological neuronal networks. “At UCSD, in the physics theory group of Henry Abarbanel, I create computational models of neuronal networks,” Armstrong told me in an email interview back in 2017, adding that her testing of the models involved a method “that is not the same as machine learning (the topic of my April Fool’s paper), but that can serve as a useful complement to it.” (Studying “impossibly-complicated” systems like atmospheres, oceans, neurons, Armstrong sought out some underlying meaningful connectivity. “That is — based on the voltage time series — infer who is talking to whom.”

So while she truly understands the use of an optimization-based framework for estimating a system’s state, she’s also exploring ways to meld scientific knowledge into other creative pursuits. (Recently the National Science Foundation even gave Armstrong the funding to create a stand-up comedy troupe of undergraduate students for “public engagement events.”) The boundaries get a little blurry sometimes. Armstrong says she was genuinely disappointed when well-meaning machine learning experts and professors had emailed her to request the complete training data sets for her junior-prom analysis, “so that they could use them in their classes.”

But alas, it was all an April Fool’s Day prank. “I was honored, and I felt badly having to let them down with the admission that the offer of supplementary materials had been part of the joke,” she told me in 2017, then adding “It’s a privilege to be in the position to make people laugh and/or think.”

Yet underlying it all is a real and sincere quest for knowledge. ” I chose artificial neural networks as the theme in order to force myself to learn how they work — and how they complement the data assimilation procedures we use in our group…” Armstrong told me.

“The opportunity to make fun of something is generally all the motivation I need to get working.”

Eve Armstrong author profile photo

Eve Armstrong

It’s not the first time she’s pulled this kind of tomfoolery. One page on her personal website is labeled simply “April 1 Findings,” where the comedy keeps coming. (“The manuscripts below are arguably my most important contributions to the science community….”) Although at least one of them appears to be based on the board game Clue.

And while the resulting paper about her junior prom was a fun spoof, there was still also one hidden grain of truth to it. “That is my high school,” Armstrong told Motherboard back in 2017, “and there was someone I never mustered the nerve to ask to the prom.” When I’d asked if she’d write more AI-based research papers on her high school crush, Armstrong joked that it depended on whether he’d drop his restraining order first.

Back in 2017, I’d argued that the satirical paper did have a larger point — “about the sophistication of our current technology, and about the messy human problems we’ll want it to solve.” As we celebrate the six-year anniversary of that legendary paper, a question began to haunt me. Whatever happened to Eve Armstrong?

I had to find out…

So it turns out that Armstrong went on to become a postdoctoral fellow for the University of Pennsylvania’s Computational Neuroscience Initiative, and from there moved on to become a real-world assistant physics professor in Manhattan at the New York Institute of Technology. By 2022 she was diligently teaching her students all about electricity and magnetism. (And to round out her schedule, Armstrong is now also a research associate at the department of astrophysics at the American Museum of Natural History.)

But Armstrong’s humor still found outlets. Her most recent paper, published last April 1, concerned “My cat Chester’s dynamical systems analysis of the laser pointer and the red dot on the wall: correlation, causation, or SARS-Cov-2 hallucination?”

Armstrong’s personal webpage describes her as “a physicist and writer with passions for theatre and science communication.”

And yet on the side, as a kind of hobby, it turns out she’s still making fun of science. But on a much, much grander scale…

Murder By Theory - Two Tales from the Ivory Tower's Dark Side - book cover - by Eve Armstrong

Murder by Theory

In an email, Armstrong told me that she’s just written and published her first book, a 290-page masterwork combining science and comedy into a pair of speculative murder mysteries titled Murder By Theory: Two Tales from the Ivory Tower’s Dark Side. Drawing on her personal experience, Armstrong’s dark novellas “cast a satirical floodlight on the gritty backstage of academia,” explains the book’s description on Amazon. (Her webpage describes it more succinctly as “a satire on science and academia.”)

In one story, mysterious notes start arriving for the residents of a small college town, where each note “explains the physics underlying a murder that the note-writer — in theory — could have wrought upon the note-receiver, but thus far has chosen not to…”

“The tutorial that accompanies this note will walk you through some simple calculations regarding momentum-conserving inelastic collisions, with applications to real-world examples to give you a feel for it. Solutions are included, but I encourage you to first attempt the exercises on your own…

“In addition, the accompanying numerical simulation code… offers you the opportunity to visualize the process by which the I-synapse term is diminished via the injection of botulinum poison. It features a handy interactive panel that allows the user to adjust the macroscopic effect on the victim. The code can be run via open-access software from any electronic device.”

One note even closes with citations from four different reference materials on the physics of pedestrian collisions. And of course, in the book, the discovery of each note is followed by a section labeled “Faculty Meeting Minutes…”

Which, suspiciously, all take place at the university’s B. Cottonfield Hall of Physical Science…

Armstrong’s author’s profile at Amazon notes that the story’s physics “includes topics that she teaches her students.” And a second story imagines the year 2033 — a dystopian future where scientists actually run the government, and a group of theoretical astrophysicists “consider it their civic duty to murder a colleague, because he is obstructing the publication of an important scientific manuscript.

“Their ensuing antics tackle the fraught peer review process and the risk of sacrificing one’s morality for career success. And they teach us that, generally, theorists are not talented murderers.”

Group Created with Sketch.
TNS owner Insight Partners is an investor in: The New Stack.
THE NEW STACK UPDATE A newsletter digest of the week’s most important stories & analyses.