Taste has become an unlikely flashpoint in Silicon Valley's AI arms race — and a source of deepening existential dread for everyone plugged into digital culture. The debate crystallized in 2026, when tech founders, investors, and engineers began invoking taste as a potential human 'moat' against AI encroachment, arguing that the capacity for genuine aesthetic judgment was one of the last things machines couldn't replicate. Critics shot back that sufficiently advanced models would eventually learn to cross that moat too, since taste is ultimately just patterns — and pattern recognition is exactly what large language models do.
The financial stakes of that bet are staggering. A multibillion-dollar AI subindustry has emerged around teaching models aesthetic judgment, with startups paying armies of human freelancers to evaluate creative output across writing, music, and visual effects. One company, Mercor, was reportedly spending over $4 million per day paying 30,000 contractors to do this work as of July — a number that ballooned to more than 100,000 contractors by November, according to a '60 Minutes' report. The moonshot goal: AI that can automate not just paralegal-style white-collar work, but abstract creative labor too.
Silicon Valley's sudden obsession with taste carries a psychological subtext. The tech industry has historically been the first elite class defined by aesthetic indifference — coders, not curators. The lone exception was Apple, where Steve Jobs and Jony Ive built a design philosophy around the work of German minimalist Dieter Rams. Ive, Apple's former industrial design chief, frames the concept simply: taste is 'fabulously subjective, and fabulously human.' That's precisely why it unsettles engineers who worship measurable data. The spectacle of Mark Zuckerberg pivoting to Prada loafers and Row knits, or OpenAI launching a streetwear drop to chase Supreme-style credibility, reflects an industry desperately trying to buy its way into cultural legitimacy.
But the anxiety around taste extends far beyond Silicon Valley boardrooms. For ordinary users, the internet has already inflicted profound damage on the experience of having preferences. René Girard's 1961 theory of 'mimetic desire' — that people want what they see others wanting — has been industrialized by algorithmic feeds that make it nearly impossible to distinguish organic discovery from manufactured desire. Algorithmic playlists surface favorite songs, TikTok influencers define 'personal style,' and 'starter pack' memes taxonomize entire subcultures into replicable codes. The fear of being an 'NPC' — an automaton following trends on autopilot — haunts users who secretly worry their own preferences are just downloaded updates rather than genuine inclinations.
The roots of this confusion run deeper than smartphones. Lifestyle marketing, coined in the 1980s, completed a transformation that began in the 1920s, when postwar production lines compelled advertisers to reframe consumption as self-expression. The counterculture that pushed back against this logic was eventually absorbed by it — anti-consumerism became its own consumer category. Today, friends sharing restaurant recommendations over group chats increasingly do so via affiliate links that earn them kickbacks. As one author frames it, we've become 'little entrepreneurs of selfhood,' and AI threatens to push that dynamic even further, with OpenAI itself acknowledging the risk of 'social deskilling' as people outsource not just tasks but thinking to chatbots.
Gist is a free Chrome extension that instantly creates AI-powered summaries and key takeaways for any article or podcast on the web.
Add to Chrome — It's Free