🐿️ Hello, Computer and How To Kill Your Keyboard

Originally published 13th January 2025

My father programmed huge mainframes with punch cards and paper tapes, but in the evening he and I would watch reruns of Star Trek and imagine that someday we might have conversational computers. We were constantly fascinated to see the crew of the Enterprise barking questions into thin air and getting useful answers from a disembodied female voice. To highlight how advanced this tech was, the scriptwriters even sent a 23rd-century engineer back in time, where he tried to talk to an early MacIntosh computer with predictably absurd results.

Perhaps surprisingly, we humans actually haven’t invented that many different ways to command our electronic servants. In the 1950s we flipped switches on the front panel; in the 60s we used keypunches and teletypes; we got alphanumeric keyboards in the 70s and pointing devices in the 80s; and – well, actually not much has happened in the 40 years since then! Sure, we use mobiles and tablets, but their touch screens are only minor advances, little more than glorified mice. Waving our hands and grabbing ghostly images a la Iron Man is possible, but hardly ready for prime time (despite me cheering from the sidelines for the Vision Pro). This paucity of innovation is why I was so surprised and pleased to hear my wife actually using the Star Trek interface last week: her ChatGPT had upgraded to Advanced Voice Mode, and just like Captain Picard, she was having a natural-sounding conversation with her phone about what it “saw” around her with the camera, a real revolution in accessibility for the world’s only blind art historian.

We’ve had voice assistants like Siri and Alexa for a few years, but they can’t actually chat – users are really just trying to guess the right magic words to say like in a 1970s text adventure (“DROP BRICK”. “BRICK LAVA”. “BLOCK LAVA”. “BUILD DAM”, aaaah!). ChatGPT is finally living up to its name as it gives us the cut and thrust of normal dialogue, complete with interruptions and intonations.

And it doesn’t matter how imperfect that interaction is – the consistent lesson of tech product development is that even the smallest crack in the barrier between users and what they want widens rapidly and inexorably into a gaping fissure. Think of Napster, which defied a billion-dollar industry and every copyright law under the sun to give users the simple ability to “type the name of a song and listen to it” – and once Fanning’s fantasy had melted university networks across the globe, not even being shut down would stop the drive for instant gratification, living on today in services like Spotify that go from title to tune at the press of a button.

Steve Jobs understood viscerally this seismic drive carrying products inexorably from interface to adoption. Of the first Laserwriter, he said that to sell it, “You don’t need to know anything about what’s in that box. All we have to do is hold this [printed page] up and say, ‘Do you want this’?” And when the contrast is so stark, between pixelated dot-matrix letters and polished fonts, or between “No, Alexa, don’t play jazz fusion, tell me if I need an umbrella today” and “Wait, tell me more about wind chill factors before you continue explaining polar vortexes”, then you’re feeling tidal forces. We’re used to people talking to themselves in public now, thanks to Bluetooth headsets and ubiquitous mobile coverage; but soon you’ll assume that person mumbling on a park bench isn’t on the phone but chatting to a robot. Is your tech organisation ready to boldly go into the voice-driven future?

This first appeared in my weekly Insanely Profitable Tech Newsletter which is received as part of the Squirrel Squadron every Monday, and was originally posted on 13th January 2025. To get my provocative thoughts and tips direct to your inbox first, sign up here: https://squirrelsquadron.com