When the news broke in late January that a Chinese start-up called DeepSeek had supposedly developed cutting-edge AI for a relative song, the story reverberated across the world’s boardrooms and social feeds. But in the corners of the internet devoted to professional naming, a question quickly followed: Have we reached peak “deep”?

True, the D-word has been a cultural touchstone for decades, and the contexts reflect the rich multitudes of the word. A yogi takes a deep breath. A quarterback throws a deep pass. We can do a deep clean or fear the deep state or have a deep thought. We might be deeply in love or deeply in debt, and sometimes both. “Deep” has associations with everything from secrecy and sexuality (Deep Throat) to interplanetary travel (“Star Trek: Deep Space Nine”).

But lately, one meaning has become trendier than the rest. Close ties to artificial intelligence have led to a surge in “deep” being used for AI-related endeavors, to the point that the word is fast becoming shorthand for “cutting-edge tech”- and is already starting to feel derivative. In 2025, “deep” is to the tech world what the plus sign (+) became a few years ago to streaming platforms such as AppleTV+, Disney+ and Paramount+.

Tech didn’t always own “deep” the way it does today. The OG “deep” is the ocean, whose depths have always promised adventure, riches and understanding to those bold enough to plumb them. It has been on trend in other ages, too: From “deep-fried” to “deep cuts” to “Deep South,” the word has fulfilled our need to describe the relationship between us and the beyond. The thread that connects all these is an exploratory instinct, a going into something, but what follows can be bad (sadness, defeat, death) or wonderful (revelation, relaxation, breading).

DeepSeek may have extra resonance in the wake of IBM’s Deep Blue, a supercomputer that bested a world chess champion in the 1990s, and Google’s DeepMind, an AI outfit founded in 2010. The earliest uses of “deep” in computing go back at least to the 1960s, according to the Oxford English Dictionary, in reference to models with multiple layers. In ensuing decades, we get variations such as “deep tech,” a term for solutions that involve substantial engineering challenges, and “deep learning,” a reference to machines imitating thinking processes.

Though the association between “deep” and “tech” is well-established, the word has lately become more widely deployed, as a simple way to suggest complex stuff that is otherwise a mouthful to explain. A search of start-up indexer Crunchbase returns hundreds of results for “deep,” many of them companies in the AI space. There’s Deepwave Digital (AI + radio frequency), Deep Instinct (AI + cybersecurity), Deepgram (AI + voices), DeepL (AI + translation), Deep Vision (AI + images and video) and DeepMap (AI + um, maps).

In the wake of DeepSeek’s splashy arrival, OpenAI announced its own offering this month: a “deep research” AI agent. The company said it was “launching deep research in ChatGPT, a new agentic capability that conducts multistep research on the internet for complex tasks.” Sounds great, but “agentic capability?” Robert Frost was right: These woods are lovely, dark and deep.

Best of all, though “deep” evokes visions of neural networks and unprecedented superskills, it makes no specific promises that a company must deliver on — much like previous buzzwords “bright” or “cyber” or “micro.” The appeal is generic, analogous to the long-lost fervor around words such as “green” (vague promises of environmental responsibility) or “smart” (vague promises that something is beneficial to use).

A general sense of ineffable progress is part of tech’s appeal, of course. Things must be oversimplified not only for commercial reasons but also for our sake, as we come along for a ride with baffling mechanisms and an uncertain destination. “Deep” may be veering into cliché, but it also works because, let’s be honest, most of us have no idea what layers are involved here.

We’re little more than ankle-deep, at the deepest.

Katy Steinmetz is a creative director at Catchword, a naming agency based in Oakland, California, and former San Francisco bureau chief for Time.