Why it’s now so hip to go analog and how AI is the antithesis of that

Cassette tapes have cracked the Billboard charts. Disposable cameras are selling out again. People are dialing friends from rotary dial landlines or screenless flip phones. Crews of self-proclaimed Luddites are rebelling in analog form against AI. These latter-day Luddites aren’t seeking to burn down server farms, but they are challenging the evangelizing of AI as a conversational and artistic partner, as propounded by AI developer fans.

While it might look like a retro fashion statement or a half-ironic aesthetic mood board, it’s actually about refusing the version of reality that AI produces, with its competently painful, painfully optimized approach to life and art that leaves no room for the surprises and flaws that define most artistic endeavors.

AI content can dazzle with the speed at which it produces a certain level of quality writing or media, but it’s a mechanical trick. While repeated experiments with prompts and iteration can push an image or text in more interesting directions, AI will tend toward a colorless, inoffensive writing style and a poreless mannequin art style. Even the videos have a weightless, empty feel when you watch them.

It’s easy to mistake polish for quality. But the best human art, or simply memorable images and turns of phrase, often sneak into view covered in metaphorical grime. But AI doesn’t do texture. It simulates, without providing insight.

Even AI’s mistakes feel wrong in a different way. It doesn’t misspell your name in an image or get your eye color wrong. It invents a disturbing new alphabet and makes your fingers appear more like a sea anemone. It doesn’t misquote a song lyric; it invents a bizarre and possibly libelous explanation for the song’s origin. These aren’t the charming errors of a human being moving fast or passionately; they’re nonsense spat out by an algorithm. It’s the uncanny valley of failure.

Artistic error beats AI perfection

In a bit of irony, it’s improvements to AI that lead to the ultra-smooth, sharp rendering and sterile, but competent writing. AI seesaws between predictably dull and chaotically useless.

The AI versus analog choice maps perfectly onto the question of making and listening to music. A computer can hit every note perfectly and simulate an acoustically flawless instrument. You might be fine with listening to its crisp fidelity from digital speakers. But left to choose, I suspect most people would favor the experience of a live performance by a virtuoso on a handmade masterpiece of a musical instrument over a manually edited recording.

Claims that AI-produced content is “spiritually empty” might be an exaggeration, but discomfort with being surrounded by only that aesthetic is probably common. And neo-Luddites itching for the scratches and bumps of handwriting and the hiss of vinyl is entirely understandable.

The machines will keep getting better, and AI could even start to mimic how humans drop notes or how you can hear the intake of breath before a voice starts singing. No matter how smooth the symphony, we’ll still crave the static.