Nieman Foundation at Harvard
HOME
          
LATEST STORY
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
ABOUT                    SUBSCRIBE
Nov. 20, 2013, 3:31 p.m.
LINK: blog.nytlabs.com  ➚   |   Posted by: Joshua Benton   |   November 20, 2013

Noah Feehan, a “Maker” at The New York Times R&D Lab, wanted to create a physical artifact that marked all the changes in Times headlines, in real time. So he built Diff,

nyt-rnd-diffa small device that monitors the internal events stream of The New York Times and prints out a summary each time an active headline is changed. As it runs, it generates a long stream of changes printed on thermal paper: text that was removed from a headline is rendered as inverted, while additions to a headline are underlined…

Of course, we were aware of and inspired by the excellent NewsDiffs project, which provides a more complete and persistent summary of changes to entire articles across several different websites. Our objective in making Diff was as much rooted in the notion of “fixing” an evanescent resource in a place and time (as NewsDiffs does) as it was a reaction to the emerging shape of “internet things” whose purpose is to transpose or transform the properties of network space onto physical space, and vice versa.

Diff found that headlines get changed roughly every five to seven minutes, unless big news was breaking.

Okay, here’s a Nieman Lab hook:

We’ve only just begun exploring the full potential of the data source for this project, which is exciting in its own right: it’s basically a near-real-time, highly-detailed stream of every event that our publishing framework sees, from the first words typed into our CMS, to an article’s publishing in its own section, to its promotion to the front page.

I unplugged Diff after a week or so of printing, and have saved the 300-odd feet of generated text for some future application. Expect to see more stream-processing tools, internet-things and interaction experiments here soon!

Feehan’s built a lot of nifty things in his young career, but for me it’ll be hard for him to top Steak Filter, in which he made a video of a steak cooking by sending the video signal through the steak as it cooked. (More cooked = less moisture = degraded signal. Now that’s exploring meatspace.)

(This is what eventually happens to the signal.)

Show tags
 
Join the 60,000 who get the freshest future-of-journalism news in our daily email.
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
Within days of visiting the pages — and without commenting on, liking, or following any of the material — Facebook’s algorithm recommended reams of other AI-generated content.
What journalists and independent creators can learn from each other
“The question is not about the topics but how you approach the topics.”
Deepfake detection improves when using algorithms that are more aware of demographic diversity
“Our research addresses deepfake detection algorithms’ fairness, rather than just attempting to balance the data. It offers a new approach to algorithm design that considers demographic fairness as a core aspect.”