The 119th Block: Fakespotting
Small lies become big kills
This week…
Let’s highlight some works done in spotting fakeness on the Internet. The methods used here are not spectacularly technical nor novel. If I make a thread of all the Internet sleuthing I have done, I would not have time to do the actual sleuthing. I could also share many a TikTok of highly motivated yet non-professional sleuths who use similar approaches to discover ‘truths’ about their current favourite online personalities or mysteries, but that’s for Garbage Day.
Of course, I caution against this obsessive “investigation culture” rampant particularly on Twitter and TikTok, but I do encourage everyone to recognise that it is not difficult to do a bit of critical fact-finding yourself and detect patterns of online deception so that if—when—you become a target of online, you are prepared.
So now, selection of top stories on my radar, a few personal recommendations, and the chart of the week.
Want to spot a deepfake video caller? Ask the suspect to turn sideways
Thomas Macaulay for TNW:
Most of the recreations were impressive when they looked straight-ahead. But once the faces rotated a full 90-degrees, the images became distorted and the spell was broken.
The team believes the defects emerge because the software uses fewer reference points to estimate lateral views of faces. This forces the algorithm unable to guess how it would look.
“Typical 2D alignment packages consider a profile view to be 50% hidden, which hinders recognition, as well as accurate training and subsequent face synthesis,” Metaphysic.ai’s Martin Anderson explained in a blog post.
For now, it works, but the technology will catch up at some point.
The case of fake IMDb credits
Prashant Baid for Pea Bee:
I was casually browsing IMDb when I landed on the page for an upcoming Ranbir Kapoor starrer movie Animal. I saw the cast details and I found a face and a name I didn’t recognise.
Finding out about this guy led me to a whole new world of how so many young Indian men from small towns are gaming the system to manufacture their own fake online clout.
An impressive work by Baid to show how with a little hustle, anyone can manufacture online notability – and your own Google knowledge panel.
In 6 minutes, Joe Hanson uncovers the true story behind viral falsehood
You’ve probably seen the viral tweet about a reusable grocery list. I’ve seen it twice in the space of months. In a Twitter thread, Joe Hanson tells you why it’s not from the 1920s as claimed:
What I read, watch, and listen to…
I’m reading how Instagram and Facebook can track anything you do on any website in their in-app browser by Felix Krause.
I’m listening to MIT TR’s In Machine We Trust on how to craft effective AI policy.
I’m watching disinfo expert Emily Dreyfuss discuss disinformation on TikTok about the war in Ukraine on CBC’s The National.
Reviews, opinion pieces and other stray links:
Permashifters want out of this reality for good by Jessica Lucas for Input.
More than monkey business: Living on common ground with our local macaques by Gwyneth Cheng for Kontinentalist.
How a minor tweak to porn sites could change the way we view gender identity by Mark Hay for Mic.
A few questions about Canada’s Artificial Intelligence and Data Act by Mardi Witzel for CIGI.
Chart of the week
The latest from Pew Research Center shows TikTok has established itself as one of the top online platforms for U.S. teens. Text report by Emily A. Vogels, Risa Gelles-Watnick and Navid Massarat is available here.
No but the comment section of the fake IMDB story is hilarious lmao