News CA

Lately: Why TikTok is so addictive, AI bubble fears and Taylor Swift’s AI slop allegations

Welcome back to Lately, The Globe’s weekly tech newsletter. If you have feedback or just want to say hello to a real-life human, send me an e-mail.

In this week’s issue:

😵‍💫 How TikTok keeps its users scrolling

🫧 Are we in an AI bubble?

🍁 OpenAI considers expansion into Canada

🤖 Taylor Swift’s AI slop allegations

SOCIAL MEDIA

How we get addicted to TikTok

According to TikTok internal documents, it takes just 260 videos – the equivalent of as little as 35 minutes – to form a habit of checking the app. But why do some users end up scrolling for upwards of nine hours a day? In a new investigation, The Washington Post collected and analyzed the viewing history of 800 TikTok users over a five-month period to see how their viewing habits evolved.

The analysis found that even occasional users were watching around 70 minutes a day, more than double the time they spent at the beginning of the experiment, while the heaviest users spent around four hours on the app per day. Occasional users would see more broadly appealing content, such as music-focused clips, on their For You page, while the power-users would be pulled into long-form “storytime” content. Both groups also opened the app more frequently as the months passed and swiped through videos quicker.

Researchers say the data trends of both groups indicate compulsive behaviour – an urge to do something even if a person knows it can be harmful in the long run. That matches what many of the users who submitted their viewing histories said too: “There are times when I know I should stop scrolling and get work done or go to sleep, but it’s so hard to stop, knowing the next swipe might bring me to a truly interesting video,” a 51-year-old operations manager told The Post. Read the full investigation here.

Are you reading this newsletter on the web or did someone forward the e-mail to you? If so, you can sign up for the Lately newsletter.

ARTIFICIAL INTELLIGENCE

Are we in an AI bubble?

Every few months, as Nvidia stocks shatter a new record and OpenAI closes another billion-dollar funding deal, the same question rears its head: Are we in an AI bubble?

This week, financial institutions are raising warning flags that the AI boom could be barreling toward a big bust. Officials at the Bank of England flagged that the “risk of a sharp market correction has increased,” while the head of the International Monetary Fund issued a similar alarm, stating that financial conditions could “turn abruptly.” There are some troubling symptoms the economists point to, such as the rapid growth of tech stocks, which now comprise about 40 per cent of the S&P 500, market valuations that are stretched beyond their worth, and enormous uncertainty around how AI will affect productivity and future jobs.

INFRASTRUCTURE

OpenAI looks at data-centre capacity in Canada

Even as some fear we could be in an AI bubble, OpenAI is eyeing Canada as a new location to build data centre capacity for training and powering its AI models. The company’s chief global affairs officer said this week that OpenAI is considering initiatives here, which could include building infrastructure or purchasing capacity in other Canadian data centres. Earlier this year, OpenAI announced it would back a large data centre in Norway and is working with the software company SAP to expand capacity in Germany.

But as Globe reporter Joe Castaldo writes, OpenAI’s desire to work with Canada comes at a time of fraught relations with the United States. There’s increasing awareness that Canada is heavily reliant on U.S. tech companies for critical services such as cloud computing and AI. This has led to a nationalistic push for more control over digital infrastructure and services, so as not to be subject to the whims of foreign governments. Read the full story here.

What else we’re reading:

The Techno Optimist’s Guide to Futureproofing Your Child (New York Magazine)

Ted Cruz picks a fight with Wikipedia, accusing platform of left-wing bias (Ars Technica)

Adult Money

JUICE WORLD

Breville Juice Fountain Cold, $270

Open this photo in gallery:

The Breville Juice Fountain Cold is a centrifugal juicer that uses cold-spin technology to reduce heat exposure, despite its powerful 850-watt motor. It processes whole fruits with minimal prep and an Italian-made mesh sieve filters the pulp. $269. Available through breville.com.Breville

I haven’t been indoctrinated into the world of fresh-pressed juice, but I appreciate that some people swear by their home juicers as a must-have kitchen gadget. The Globe recently did a deep dive into home juicers, breaking down the best buys for all price ranges. This Breville juicer comes highly recommended: it’s centrifugal, which means it has fast-spinning blades and high-speed mechanisms to separate juice from pulp, and can process whole fruits with minimal prep.

Culture radar

AL SLOP

Did Taylor Swift use generative AI to market The Life of a Showgirl?

Open this photo in gallery:

Screenshot from a Taylor Swift video of a decorated piano with odd shapes and lack of detail that suggests it was made with generative AI.Supplied

Over the weekend, Taylor Swift released a series of promotional videos for her new album The Life of a Showgirl. The videos were filled with Easter eggs – the signature hidden clues and inside jokes that Swift tucks away for her fans to find and decode. But Swifties quickly noticed that something felt a bit off with these videos. There was a squirrel with an odd appendage sprouting from its back, a treadmill with gibberish buttons, inconsistent fonts, and hallucinated hands. It all looked a lot like AI.

“The discovery that AI might be at the centre of Swift’s latest claim to uberwealth has left me feeling more than mere disappointment,” writes Aisling Murphy, The Globe’s theatre critic and Swiftie-in-chief. “It’s something darker, sharper – an abandonment of the lighthearted games that once made being a Swiftie so special.” Indeed, Murphy notes that Swift’s seeming pivot to AI also feels hypocritical. Less than a year ago, the singer spoke out against the technology in the context of deepfakes. “The Swift who may have deliberately used AI to fog her own work is a Swift I don’t recognize,” says Murphy. Read the full story here.

More tech and telecom news

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button