Want to hear something eerie? An AI Frank Sinatra song, seriously. Turns out if you feed a lot of Frank into a computer and press a few buttons it can “create” a new track. But just because it’s possible doesn’t mean you should.
It’s technically perfect and it sounds like him, sort of. Except it doesn’t where it matters. It’s just a bit “off” but you can’t quite put your finger on why. Turns out the sum of all those gigabytes isn’t Frank Sinatra, however much they claim to be.
Truth is, for now, the human element, the pause, the pronunciation of a word, the vibrato of a note, the “mistakes” are really hard to reproduce using AI. And as humans we are used to those human touches. We like them, recognise them and we notice their absence. In a nutshell it’s what brings emotion to the music and connects us to the artist.
The same is true of the written word. The looming presence of reams of computer generated literature is frightening on many levels. Will it make writers a dying breed? Or create increasingly homogenised content? Words consigned for eternity to ever decreasing circles of blandness.
I don’t believe it will. 1s and 0s are never original. We love novelty, variety, innovation, the wrong note. No doubt AI is amazing, just as long as it leaves my records and books alone.