AI & the Literary Imagination
AI & the Literary Imagination
Could the next great American novel be written by a machine?
View this email in your browser
Will the next Shakespeare be powered by algorithms?
We hope this finds you safe and well. This week, we’re offering a fleeting respite from the grave reality of the global public health pandemic of COVID-19.
Many people use natural language processing tools rather regularly. Perhaps most ubiquitous is sentence completion tools in email service providers. But how does this translate to less literal uses of language? As fans of classic, modern and experimental writers and the written word in its many forms, we are looking towards a non-digital form of entertainment: Creative Writing.
We’ll explore machine generated novels and poems currently in the world, and ask a couple questions: “Could the next great American novel be authored by a machine?” and ultimately, “Is it any good?”
We’re also considering whether AI really is a technology or something else and learning how Square co-founder Jim McKelvey defines innovative companies. We’ll conclude with a note on Asimov’s Laws for robot writing.
We are also highlighting tech to support mindful and meditation practices from our two-part series on tech and mental health. And, for those who may be struggling with a bumpy transition to remote work, we have some ideas on workflow optimization. We hope that the tips and resources in these previous issues may be of use during these extraordinary and challenging times.
We hope you’ve been enjoying this newsletter and would love any feedback (erica@nycmedialab.org). We wish you and your community safety, calm and solidarity as we support each other. Thank you again for reading.
All best,
Erica Matsumoto
As the world continues to grapple with COVID-19, we, at least, are in desperate need of some relatively more lighthearted content. In the new era of social distancing to keep ourselves and others healthy, intellectually escaping via reading can be a source of comfort.
On a fun, related note: if you find that the classic novels you’re reading right now don’t resonate in the time of social distancing, check out these first lines of 10 classic novels as rewritten for social distancing.
Let’s explore the strange world of AI literature.
AI LITERATURE & POETRY
For a taste of what fully AI-produced literature could look like, we can first look to 1 The Road, the first-ever AI-written novel. The novel is about a novelist’s journey across the U.S. attempting to emulate Jack Kerouac (the mission: go out on the road and find something essential in the experience to write about).
While the novel’s opening line (“It was nine seventeen in the morning, and the house was heavy.”) works, much of it doesn’t make sense. For example, the locations it describes are based on Foursquare data, so the text randomly quotes longitude and latitude coordinates verbatim.
Ross Goodwin, the self-described “gonzo data scientist” and former ghostwriter for Barack Obama who created the AI that wrote 1 The Road, describes that AI as “not quite human level,” but rather “more like an insect brain that’s learned to write.” That’s perhaps not a ringing endorsement of AI’s ability to produce Kerouac-level prose.
Goodwin isn’t the only person tinkering with AI for literature, either: NYU professor Allison Parrish is a poet and programmer who uses AI to produce poems, chatbots that mimic people or things, and more. Check out Parrish’s portfolio for links to fascinating work. In particular, Compasses is a mind-bending exploration of what, precisely, constitutes “poetry” in an AI’s imagination
Parrish isn’t the first to use AI to write poetry: in 2018, a team of Microsoft and Kyoto University researchers developed a poet AI good enough to trick online judges into believing its output was produced by a human.
In yet another application of AI to the world of poetry, Chaim Gluck, an independent machine learning consultant, built a machine learning model that uses the differences in word and style choices between centuries to determine when poems were written to a surprisingly high degree of accuracy.
To play around with AI as a substitute for producing human language (and to cure the monotony of working from home if you’re starting to bounce off the walls), check out Talk to Transformer.
AI TEXTBOOKS
While AI may not yet be up to the task of writing flowing, human-like prose novels, it does seem up to the task of writing textbooks. In April 2019, academic publisher Springer Nature unveiled what it claimed was the first-ever research book generated using machine learning, Lithium-Ion Batteries: A Machine-Generated Summary of Current Research.
The book is a summary of peer-reviewed papers on, you guessed it, lithium-ion batteries. It has quotes, hyperlinks to cited work, and automatically-generated references. In the intro, Springer Nature editor Henning Schoenenberger says books like this have the potential to start “a new era in scientific publishing” by automating the tedious work of scanning, summarizing, and citing existing literature.
Science textbooks aren’t the only domain in which AI proves its worth as a writer of rote text: the Associated Press uses machine learning to create summaries of football matches, earthquakes, and financial news. In short, it uses AI to write about topics where creativity is a shortcoming rather than an asset. WHAT DOES THIS MEAN FOR WRITERS?
With the advances being made in the realm of AI literature and AI textbooks, one might wonder, “Is this the end for human writers?” Could there come a time when flesh and blood writers are no longer needed to write stories that speak to human emotions and launch movie adaptations?
Thankfully it’s unlikely that it’ll come any time soon. Canadian novelist, essayist and futurist Karl Schroder predicts, “[Eventually] AI will be able to create a book worthy of the name, but certainly not in its current form. These will be different kinds of machines, which we have not yet thought up. Today’s computers do not produce meaning, and human intervention is always necessary in the creative process, even if technological devices are becoming more refined and approaching human capabilities.”
From U.S. to Europe to China to Russia, renewables are now cheaper than coal In all the world’s major markets, it’s now cheaper to invest in renewable energy over coal. By 2030, the Carbon Tracker Initiative estimates that new renewables will be cheaper than existing coal plants in all major markets.
Source: Fast Company 2 min read AI is an Ideology, Not a Technology Although we typically think of AI as a technology, there’s some evidence supporting the idea that it is, in fact, an ideology. Specifically, it is an ideology that relies on an “ignore the guy powering the curtain” mirage to privilege technological work (and those who perform technological work) over all other forms of work and workers. 5 min read Square’s Cofounder on Discovering — and Defending — Innovations
Square co-founder Jim McKelvey argues that most companies that believe they’re innovators are actually just copycats. He argues that true innovation requires “fearlessly exploring novel solutions and dramatically expanding markets,” which most startups don’t actually do, even if they believe they do.
25 min read This Month in Business History
March 1942: American writer Isaac Asimov proposes three rules about writing robots in his short story “Runaround”
Asimov’s three rules are:
- A robot may not injure a human being or, through inaction, allow a human being to come to harm.
- A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
- A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.
“Runaround,” like many of Asimov’s Robot stories, centers its conflict around a problem in the application of the Three Laws of Robotics. In the story, the robot finds it impossible to obey both the Second and Third Laws at the same time; this freezes it into a loop of repetitive behavior.
Finally, for a bit of levity, please enjoy this fun clip (click through to play on Twitter):
This email was sent to <<Email Address>>
why did I get this? unsubscribe from this list update subscription preferences
NYC Media Lab · 370 Jay Street, 3rd floor · Brooklyn, New York 11201 · USA
