From Bloomberg:
Botnik is creating an unusual predictive keyboard—suggesting words based on what’s been typed—to generate everything from scripts for new episodes of Seinfeld to funny Valentine’s Day recipes. The results are by design weird as hell.
. . . .
Art created by artificial intelligence has become a reliable success in the finicky world of viral content, resulting in everything from eerie cat drawings to dadaist punk music. Botnik’s interactive keyboards let anyone create surreal rearrangements of familiar words.
. . . .
At the New Yorker, Mankoff created the caption contest, spawning a huge data set mined by Google. This piqued his interest in AI, and he got in touch with Brew, who’d been exploring the topic by sending texts on the iPhone’s predictive keyboard. Botnik made its debut in 2016, then landed a $100,000 contract from Amazon.com Inc. to help make its Alexa AI assistant sound more human.
. . . .
Ultimately, Brew looks at the content created by the broader Botnik community as advertisements for the real product: the virtual keyboards themselves, which roughly 1,000 people per day play around on. Two full-time programmers have been working on a broader platform evolved from the keyboards, to be unveiled this summer. Eventually, Brew and Mankoff hope to charge for access to the platform.
Link to the rest at Bloomberg
I plugged a paragraph of my own story into the text and let it continue writing it.
“Evil mermen don’t want her life until Marielle meets mortal. For eight years he thought she deserved doozy deadbeats mccracken werecougars.”
I’m no longer worried Botnik will be taking my job next week. But I’m also incredibly intrigued to learn more about “doozy deadbeats mccracken werecougars.”
Anyone else here ever have the auto-complete/auto-correct suggest the ‘wrong’ word?
Just as an error can jog a reader out of a story, this unhelpful ‘help’ could knock a writer out of their ‘grove’ …