I learned so much while putting together Season Two of the Technically Optimistic podcast that I want to share some of my findings. Last week, I wrote about Episode One’s key points — namely surveillance capitalism and what exactly is data (seeing as it’s the season’s theme). Listen to hear great interviews with Amba Kak, executive director of the AI Now Institute; Chris Wiggins, head of data science at the New York Times and co-author of How Data Happened; media scholar, tech writer, and internet activist Ethan Zuckerman; engineer and inventor of the cookie Lou Montulli; and Harvard professor Jonathan Zittrain.
This week, I was honored to have another round of incredible guests, including Flipboard co-founder Mike McCue, Jonathan Zittrain, professor of public policy and computer science at Harvard (and my former professor); and Meredith Whittaker, the president of the Signal Foundation, which created the fully encrypted messaging app SIgnal. We were also extremely fortunate to be able to sit down with Frances Haugen, who sent me down a particular data rabbit hole that I’d like to share with you.
Haugen is a data scientist who spent 15 years in the tech world, working at companies like Google, Yelp, Pinterest and, starting in 2019, Facebook. You probably know her as the Facebook Whistleblower. That’s because in 2021, she shared documents with the Wall Street Journal showing that the company had a casual attitude, at best, toward stopping the spread of disinformation; that there’d been a rollback in policies aimed at protecting civic safety; and that the company was aware of serious mental health risks to younger users of Instagram, owned by Facebook — which actually changed its name to Meta a few weeks later.
Digging Through My Data
It was incredible to get her insights on how companies collect data on us. But the biggest takeaway came when Frances asked me to do something simple that I’d never considered: Look at the data that a social media platform has on me. I chose Instagram. I was shocked, to say the least.
You can do this on any of the Meta platforms, and I highly recommend that you do it in order to get a visceral understanding of the data that Meta is collecting on you. Just go into the data export tool in Facebook or Instagram and request your data. You will be emailed a link to download a massive ZIP file — several gigabytes. While I was sifting through it, I found everything from fundraisers I’ve attended to purchases I’ve made.
Here’s the transcript of Frances walking me through it:
RAFFI: On the very bottom, there’s Instagram ads and businesses. Advertisers can use or upload a list of information that we can match to your profile. Oh, so what they’re saying is like these advertisers have a list of types of people they want to reach and Instagram’s like, “okay we can reach those people for you” and then they do it. It’s a crap ton! It’s, like, not a small list. There are easily hundreds. Travelers choice, Bike Monkey, Burberry, I mean literally this list — I haven’t even gone down five percent of the list. There are some names like The Container Store or Canon, like huge brand names, and then there are also things I’ve never heard of like…Epic Gardening. This list is ridonculous.
FRANCES: And one of the things that would fit within there is the topicality. These are kind of like the subjects of what they think you're interested in.
RAFFI: Ah, yeah I see it. “Add topics that are relevant to you.” Oh, this is awesome. Visual arts, which is possibly true, science fiction, gym workouts, which is also possibly true, fashion, and then hair care. I wonder if beards are involved in haircare.
FRANCES: And so that topic model ends up influencing what you get to see or don't get to see, but you don't necessarily know that Facebook has put you in these buckets, and that the world that you get to witness is influenced by them.
This simple, if scary, exercise gave me incredible insight into how I’m seen, and, more importantly, what I’m shown. Am I really my data?
Please try it and let me know what you find out about yourself. You can email me at us@technicallyoptimistic.com.
We’ll deeper into this in the coming weeks. I hope you will subscribe and listen along.
Worth the Read
Meta AI rolled out last week, and the creepy data creep has already begun, as this tweet shows.
Why is it that only tech companies and drug dealers call their customers users? A thought-provoking read.
The TikTok legislation passed with record speed: The first tech bill that’s made it across the finish line in years. Will other bills be able to ride on its luck?
Also via the NY Times, Saudi Arabia is working out how to become an AI superpower.
Google is (slowly) getting rid of cookies. Here’s what that looks like.