TR.BIZ: 2.12.2026

From a new J.Crew collab to a Threads AI assistant, this is your late-mid-week check-in 💫

Welcome to The Trend Report: Business Edition™, a midweek look at top stories, trends, and more of what’s happening online and off by Kyle of The Trend Report™. Today, we’re diving into a great J.Crew concept, a Threads algo tweak, conversation hearts for 2026, a theory on how Boy Throb is slaughtering Simon Cowell’s new band, and why AI music is going to kill Spotify.

---

👠 Aesthetically Pleasing: Will the J.Crew collabs save J.Crew?

Brand collabs are nothing new, instead a retail expression that has been folded into the modern shopping experience alongside things like drops. Much of the product and shopping world now is deeply uninteresting as a result, which is why everything but fashion (Food! Decor! Music!) “is fashion” now. Imagine my surprise then that J.Crew — the lovable American brand that has been wobbly for some time — actually pulled off an interesting multi-layered collaboration: to celebrate their ongoing obsession with rollneck collar sweaters, the brand has enlisted five “new” American brands to remix the look, including Buci and Patrick Taylor but also Collina Strada (!), Tanner Fletcher (!!), and Eckhaus Latta (!!!!!). “These designers are helping to shape the next chapter of American fashion, and it’s exciting to see this signature style come to life through these fresh perspectives,” Olympia Gayot, creative director of J.Crew, said of the collab. This is major considering someone like Eckhaus Latta repeatedly is the most exciting thing in American fashion, turning this mini-collection into a dog whistle for those of us (Me.) who are interested in contemporary design and have been rooting for — but have been deeply bored by — the preppy mainstay. We’ll see if this sells but, regardless, it’s a smarter step to lead with tastemakers than to opt for celebrities or Olympics shit (or the tainted magic that is over-reliance upon sales).

📲 Tech Talk, I: Will an AI butler make social better?

I got a PR tip that Threads has officially launched a new feature called Dear Algo, an “AI-powered feature that gives you more control over what you see by letting you tell Threads what’s important to you in the moment.” Sure? This concept has been rumored for months but now is available to everyone. It works by adding “dear algo” to a post plus what you want to see, which then guides your feed toward more or less of a subject, an adjustment that can last for up to three days. Requests are public and in feed — so you can play with fire with your algorithmic desires in front of a crowd — all suggesting how we’re reaching the limits of “for you” style algorithms: the problem of unruly and uncurated feeds created by tech companies is now being solved by another feature by tech companies. A watchman to watch the watchman, as we stray further from anything straightforward. Stay tuned if this takes off, or if a platform like Twitter copies it.

📲 Tech Talk, II: Algorithmic music keeps us up at night

KRF NOTE: This piece by Olivia Choi came up in a conversation recently and was…fascinating! Excuse the length as my angel bb cooks up some big brained Spotify thoughts.

Spotify is my escape. The premium subscription? One of the best investments I’ve made in my life. I’m a big fan of daylist and Spotify Radio too, where the app’s algorithm helps me find new songs based on past preferences. However, ever since the data theft in December, I learned a lot more about how AI might impact music consumption than I ever wanted — and we need to talk about it.

First, I learned that what I most love about Spotify — their recommendation feature — may not actually be a good thing: this algorithm creates a feedback loop focused on retention, enabling what researchers call “taste tautology,” meaning listeners are siloed to musically repetitive bubbles. For example: when I listen to “the cutest pair” by Regina Song, a mellow-pop song by female Asian artist, the algorithm suggests “cloud 11” by thủy, another mellow-poppy song by another female Asian artist. Unless I force myself to explore different genres of music, algorithms like this constantly reinforce past preferences, gradually narrowing my taste down to a predictable mathematical average of what I’ve already been listening to instead of what I will evolve to like. Including myself, many people have pointed out Spotify’s recommendation feature as an edge differentiating them from competitors like Apple Music, a services that mixes algorithmic and human curation. This poses a liability: Spotify may be preferred but listeners are obvious more susceptible to such a “feedback loop.”

To make matters more icky, AI is now inspired by such repetitive algorithms — which is also tricking the algorithm too. Leaked data says that music uploads to Spotify are growing rapidly, jumping from around eight million albums in 2023 to over ten million in 2024. This spike is largely blamed on AI and procedurally generated content. As we know, AI is trained on existing hits, producing music that adheres to the most common chord progressions and structures, often mimicking popular styles just enough to “trick” the algorithm into placing them on major playlists like Release Radar. Are you familiar with Velvet Sundown? Or Xania Monet? Well, neither are real despite the first having a major viral breakout last summer and the second charting on Billboard last fall. What about this song? I really enjoyed it — but I still can’t figure out if this singer is human or not. And I may never know!