Wendy’s, Google, and Instagram

Three times the media missed important context in the last week.

By Brad Berens

I read a lot of High Quality News and avoid the Low Quality variety (see the helpful way that Ad Fontes Media defines these categories). Even topflight news can be so focused on short term events that they forget to ask context questions, by which I mean the questions that ought to follow who, what, where, when, how, and why. Those questions are so what? who cares? and says who?* These are the questions that dig deeper and look for underlying causes.

Image created by DALL-E after I shared the text of this article

Over the last week, I ran into three frustrating examples of missed context. (There were more, but these are the three I’ll dig into.)

Where Wendy’s went wildly wrong

In an earnings call, new Wendy’s CEO Kirk Tanner talked about the fast-food chain’s plans to experiment with “dynamic pricing” (H/T Warren Pickett). Reporters then compared this to Uber’s much-loathed surge pricing (where rides get more expensive during busy times). A huge backlash ensued. Even Senator Elizabeth Warren took to X to decry this as price gouging. A day later, Wendy’s was in full crisis mode, denying the media account ($) and — in the first missing story – accurately pointing out that it had never used the word “surge.”  The comparison to Uber was entirely the creation of journalists.

The second missing story: There’s nothing wrong with using financial incentives to get diners to show up at less-busy times when it’s convenient for a restaurant: it’s just an early bird special. If Tanner had referred to “dynamic discounts” rather than dynamic pricing this would never have become a media story.

We know from Behavioral Economics that people are twice as likely to feel bad about losing something than they are about gaining something. With loss aversion, you feel twice as bad about $5 dropping out of your pocket than you feel good about finding $5 on the street. “Dynamic pricing” sounds like a loss: sometimes you’ll have to pay more; “dynamic discounts” sound like a gain: sometimes you’ll get to pay less.

How a business frames its actions is important.

Google’s Gemini gone awry

Alphabet, Google’s parent company, lost a whopping $90B in market cap last week after its new ChatGPT rival, Gemini, failed a series of predictable stress tests from users kicking the tires. Gemini’s image generator, asked to create a picture of WWII era Nazi soldiers, did so with people of color as those soldiers ($). While this is a historically inaccurate and somewhat offensive result, it’s not a surprise: Google had programmed Gemini to avoid a well-known algorithmic bias problem with AI, and Gemini overcompensated. The company temporarily took Gemini’s image generating ability’s away in response.

The missing stories: We’ve seen this movie before. For example, in 2016 Microsoft released a Twitter bot called Tay, but the company took Tay down a day later because mischievous or malignant users taught Tay to say racist and misogynistic things. Microsoft suffered no longterm disadvantage because of this kerfuffle.

Likewise, this story won’t hurt Google: it’s a PR speed bump rather than a catastrophe. Although $90B is a staggering amount of money for a person, Alphabet’s market cap at the time I’m writing this is $1.7T, so the temporary loss is around five percent. This is not great, but it’s not a big story since the value of the company has more than doubled in the last five years (click the 5Y option).

Moreover, the AI wars won’t be won and lost because of bots like ChatGPT and Gemini. Generative AI is important, but it’s only one part of the AI story. Google’s ability to integrate different forms of Machine Learning into search, Gmail, YouTube, Assistant, Docs, Drive, Sheets, Calendar, Slides, Classroom, Chrome, Android, and everything else it offers—mostly for free while Microsoft charges annual fees for 365—is the real story.

Google with AI will increasingly personalize all those services, creating switching costs and generally being so helpful that you don’t want to look for alternatives. In comparison, how important is a bot’s ability to create historically accurate images?

The media’s focus on Gemini’s ham-handed responses also misses a genuine threat to Google, Perplexity. This is an AI powered “knowledge engine” that organizes information and answers questions. It shares its sources (unlike ChatGPT), so if it does hallucinate the user can easily see where the AI went off the rails. Jeremiah Owyang was the first person to tell me about Perplexity, saying that his use of Google had plummeted, and subsequently Kevin Roose had a similar story in The New York Times ($).

In the last few months, Perplexity has become my go-to when I have a question. I even pay $20/month for the Pro version. Unlike Alphabet, Perplexity doesn’t yet have a reassuring business model (Google makes its money through advertising), but if large numbers of users reduce their Googling, that’s bad for the company.

Instagram’s “mom-run” accounts for their kids… and pedophiles

Trigger Warning: This is a horrifying story from the front page of last week’s Sunday New York Times about Meta’s depraved indifference to the sexualization of underage girls on Instagram ($).

How it works: moms set up Instagram accounts to help their very young, underage daughters (as young as five years old) become influencers (which is weird). These accounts then serve as an all-you-can-eat buffet for pedophiles, who have organized resources to share information with each other.

Everybody is wrong here. The moms are wrong to set up the accounts in the first place, and then wrong again when they don’t shut down the accounts the moment a pedophile starts asking for sexy pictures of their daughters. The girls are mistaken to think that dressing up for strangers is a reasonable thing to do, although these are children, so they bear no blame. (And where are the dads in all this?) The pedophiles are, of course, terrifyingly wrong.

Meta is the biggest culprit. The company’s response to being a platform for pedophiles is to blame the parents, but it doesn’t acknowledge that it makes money both from general engagement and also from the paid subscriptions it runs for Instagram influencers. The company also limits how many accounts a user can block in a given day:

“I remember being told, like, I’ve reached my limit,” said a mother of two dancers in Arizona who declined to be named. “Like what? I reached my limit of pedophiles for today. OK, great.”

The missing story: there are other singers in this operatic chorus of wrongness that the article never mentions: the local, state, and federal governments that do nothing to stop social media companies from making things easy for pedophiles, most particularly the federal government.

Section 230 is part of the Telecommunications Act of 1996; it absolved internet companies of liability around things that users put on websites (which I’ve explored previously). While Section 230 was part of the engine that drove the growth of the internet and transformed life as we know it, that doesn’t mean that it should stick around, particularly when the social harms now outweigh the benefits.

If the federal government sunset or reset the protections of Section 230 and made Meta (and its peers) liable for the content that users put on its platforms, then you would see the company act immediately to contain this and similar problems of abuse, hate speech, misinformation, and disinformation.

I admire the hard work and passion that goes into High Quality News, and I acknowledge that this is a crisis moment for news organizations around the world as the revenue models for journalism are collapsing in real time. However, even the highest quality news can always do better, dig deeper, and help us to see underlying context.


Brad Berens is the Center’s strategic advisor and a senior research fellow. He is principal at Big Digital Idea Consulting. You can learn more about Brad at www.bradberens.com, follow him on Post and/or LinkedIn, and subscribe to his weekly newsletter (only some of his columns are syndicated here).


* I’ve adapted these context questions from Jonathan Klarfeld, a longtime Boston University journalism professor who died in 2018. His student, my lifelong friend David Daniel and himself a career long journalist, passed these insightful questions along to me.


See all columns from the Center.

March 8, 2024