This week…
I built a couple of pieces of large furniture. There were several trips to the furniture store and back. My body aches. WHO declares monkeypox a global health emergency. The US records its first case of polio since 2013. What a week. And:
Now here’s a selection of top stories on my radar, a few personal recommendations, and the chart of the week.
The Great Fiction of AI
Independent writers turn to AI to meet Amazon’s demands to publish books as quickly as possible, writes Josh Dzieza for The Verge:
[Jennifer] Lepp, who writes under the pen name Leanne Leeds in the “paranormal cozy mystery” subgenre, allots herself precisely 49 days to write and self-edit a book. This pace, she said, is just on the cusp of being unsustainably slow. She once surveyed her mailing list to ask how long readers would wait between books before abandoning her for another writer. The average was four months. Writer’s block is a luxury she can’t afford, which is why as soon as she heard about an artificial intelligence tool designed to break through it, she started beseeching its developers on Twitter for access to the beta test.
The tool was called Sudowrite. Designed by developers turned sci-fi authors Amit Gupta and James Yu, it’s one of many AI writing programs built on OpenAI’s language model GPT-3 that have launched since it was opened to developers last year. But where most of these tools are meant to write company emails and marketing copy, Sudowrite is designed for fiction writers. Authors paste what they’ve written into a soothing sunset-colored interface, select some words, and have the AI rewrite them in an ominous tone, or with more inner conflict, or propose a plot twist, or generate descriptions in every sense plus metaphor.
Eager to see what it could do, Lepp selected a 500-word chunk of her novel, a climactic confrontation in a swamp between the detective witch and a band of pixies, and pasted it into the program. Highlighting one of the pixies, named Nutmeg, she clicked “describe.”
“Nutmeg’s hair is red, but her bright green eyes show that she has more in common with creatures of the night than with day,” the program returned.
Lepp was impressed. “Holy crap,” she tweeted. Not only had Sudowrite picked up that the scene Lepp had pasted took place at night but it had also gleaned that Nutmeg was a pixie and that Lepp’s pixies have brightly colored hair.
Commercial image-generating AI raises all sorts of thorny legal issues
Now available in Beta, DALL-E 2 is ready to sell. Kyle Wiggers writes for TechCrunch:
This week, OpenAI granted users of its image-generating AI system, DALL-E 2, the right to use their generations for commercial projects, like illustrations for children’s books and art for newsletters. The move makes sense, given OpenAI’s own commercial aims — the policy change coincided with the launch of the company’s paid plans for DALL-E 2. But it raises questions about the legal implications of AI like DALL-E 2, trained on public images around the web, and their potential to infringe on existing copyrights.
DALL-E 2 “trained” on approximately 650 million image-text pairs scraped from the internet, learning from that dataset the relationships between images and the words used to describe them. But while OpenAI filtered out images for specific content (e.g. pornography and duplicates) and implemented additional filters at the API level, for example for prominent public figures, the company admits that the system can sometimes create works that include trademarked logos or characters.
Explained: How to tell if artificial intelligence is working the way we want it to
And now for an explainer/refresher. Adam Zewe for MIT News:
About a decade ago, deep-learning models started achieving superhuman results on all sorts of tasks, from beating world-champion board game players to outperforming doctors at diagnosing breast cancer.
These powerful deep-learning models are usually based on artificial neural networks, which were first proposed in the 1940s and have become a popular type of machine learning. A computer learns to process data using layers of interconnected nodes, or neurons, that mimic the human brain.
As the field of machine learning has grown, artificial neural networks have grown along with it.
Deep-learning models are now often composed of millions or billions of interconnected nodes in many layers that are trained to perform detection or classification tasks using vast amounts of data. But because the models are so enormously complex, even the researchers who design them don’t fully understand how they work. This makes it hard to know whether they are working correctly.
For instance, maybe a model designed to help physicians diagnose patients correctly predicted that a skin lesion was cancerous, but it did so by focusing on an unrelated mark that happens to frequently occur when there is cancerous tissue in a photo, rather than on the cancerous tissue itself. This is known as a spurious correlation. The model gets the prediction right, but it does so for the wrong reason. In a real clinical setting where the mark does not appear on cancer-positive images, it could result in missed diagnoses.
With so much uncertainty swirling around these so-called “black-box” models, how can one unravel what’s going on inside the box?
What I read, watch, and listen to…
I’m reading how Shannon Curley became one of the Blue Jays’ most influential voices by David Singh for SportsNet.
I’m watching world No. 7 men’s singles tennis player Andrey Rublev and Russia’s No. 1 women’s singles tennis player Daria Kasatkina in a revealing interview with vlogger Vitya Kravchenko. They speak out against the war in Ukraine and discuss Kasatkina’s sexuality (she dates Russian-Estonian Olympic figure skater Natalia Zabiiako). I don’t think they will compete for Russia for long.
I’m also watching The Undeclared War. Read a review in the next section.
I’m listening to this. It’s the music that keeps playing on The Undeclared War. I don’t understand it but it’s an earworm.
Reviews, opinion pieces and other stray links:
What only musicians and translators know is how to take the cerebral element and turn it back into emotion for a new audience, writes Jessica Sequeira on LitHub.
If language began with gestures around a campfire and secret signals on hunts, why did speech come to dominate communication? asks Kensy Cooperrider on Aeon.
How realistic is Peter Kosminsky’s The Undeclared War? Very, according to one of the UK’s top digital intelligence experts, Alan Woodward, writing for The Guardian.
Chart of the week
Instagram, TikTok and Youtube are the top three news sources for UK teens, according to Ofcom’s news consumption report.
So... how do we get our hands on Sudowrite?