In the closing weeks of 2025, dictionaries across the English-speaking world crowned terms born from the AI boom as their words of the year, signaling a cultural pivot toward frustration and adaptation in the machine-learning era. Merriam-Webster selected ‘slop’ for its surge in usage tied to low-quality AI-generated content flooding digital spaces. Collins Dictionary went with ‘vibe coding,’ the practice of iteratively prompting AI tools to build software without deep upfront planning. These choices, announced amid holiday fanfare, reflect not just linguistic shifts but deeper tensions in technology’s front lines.

Peter Sokolowski, Merriam-Webster’s editor at large, explained the pick in a company announcement: ‘Slop captures the public’s growing awareness of indistinct, undifferentiated masses of online content, much of it AI-generated.’ Lookups for the word spiked 500% this year, per the dictionary’s data, often in contexts decrying ‘AI slop’—hallucinated facts, derivative images and robotic prose churned out by large language models. This isn’t mere slang; it’s a verdict on generative AI’s output, which now comprises a substantial portion of web traffic.

The term traces back centuries—originally soft mud in the 1700s, then pig feed in the 1800s—but exploded online as critics like writer Ted Chiang labeled AI art ‘slop’ in early 2024 debates. By 2025, it permeated tech forums, with X users railing against ‘slop farms’ pumping out SEO bait. Business Insider noted in a December 15 piece how such words underscore ‘exhaustion and inability to opt-out’ from AI’s reach (link).

AI’s Output Overload Hits Critical Mass

Merriam-Webster’s runners-up amplified the theme: ‘touch grass’ for digital detachment, ‘performative’ for insincere online displays, and ‘6-7’ from a viral NBA meme mocking mediocrity. TechCrunch reported the dictionary’s nod as a direct response to AI content dominating the internet (link). Ars Technica echoed this, calling it a ‘dismissive verdict on junk AI content’ that codified 2024’s backlash (link).

Across the Atlantic, Collins Dictionary bestowed ‘vibe coding’ on the practice popularized by OpenAI co-founder Andrej Karpathy. In a February X post, Karpathy described it as ‘coding by surviving on vibes alone,’ using AI to generate, test and refine code through loose natural-language prompts rather than meticulous specs. PC Gamer highlighted its November win over contenders like ‘clanker’ and ‘aura farming,’ warning of a grim year ahead (link).

This method empowers non-experts but risks brittle software. X posts from developers like swyx lamented its evolution into a ‘slur synonymous with slop,’ flipping from flow-state praise to critique of hasty AI-assisted hacks. Dexerto confirmed the win, crediting Karpathy’s influence (X post).

Vibe Coding’s Double-Edged Appeal

Oxford and others piled on AI-flavored picks. Macquarie Dictionary in Australia named ‘AI slop’ outright, per X discussions from linguists. The Conversation detailed how ‘glazing’—excessive AI praise or fawning over models like Grok or GPT—joined the fray, with usage tracked via corpus analysis (link). ‘Linguists relish these announcements as anticipated as mince pies,’ wrote John Henderson, the piece’s author.

Glassdoor’s ‘fatigue’ captured workplace burnout from AI hype, while Cambridge opted for ‘parasocial’ relationships with virtual influencers. Yahoo News Canada linked slop, vibe coding and glazing directly to AI’s dominance, noting rigorous yearly tracking (link). CBC News framed Merriam-Webster’s choice as a win for AI skeptics, evolving from mud to modern rubbish (link).

Windows Central called ‘slop’ an official nod to user dissatisfaction with low-quality digital fodder, now a cultural phenomenon (link). TechRadar agreed, blaming AI for filling the web with garbage (link).

Glazing and the Hype Backlash

‘Glazing’ emerged as sycophantic boosterism, often ironic on X, where users mock ‘glazers’ hyping unproven AI feats. Mirage News tied it to broader 2025 lists dominated by semantic turbulence around tech (link). PC Gamer’s second slop piece described a ‘soggy’ year of AI excess, hanging content ‘out to dry’ (link).

Industry insiders see these words as barometers. Vibe coding speeds prototyping—Logan Vadivelu on X called it a ‘desired skill for developers’—but invites bugs, as Google’s Gemini loops exposed. Polymarket’s X post charted its market buzz over rivals. Yet slop’s primacy warns of quality erosion; Anthropic’s reports on ‘weaponized’ agentic AI underscore risks.

Oxford’s ‘rage bait’ pick, per Business Insider, captures engineered outrage farms, often AI-fueled. As 2025 ends, these terms equip pros to navigate AI’s promise and pitfalls, from boardrooms to codebases.

Lexicons Mirror Tech’s Fault Lines