Opinion | Have we already lost control of AI? Musings on Meta and Moore’s Law, facts about frogs in saucepans
- Doctor Who’s Daleks had it right, it seems: resistance is useless as AI quickly expands its reach
Among the boosters and sceptics in the world of artificial intelligence, I’m staying right here on the fence. My booster shoulder points towards nerdy fun and the chance to argue with AI about its artificiality (not artificial: it’s human data, pillaged and processed) and intelligence (alien in a human suit).
It’s useful navigating from place to place, and it can advise on issues ranging from recipes, etiquette, use of space – including detailed discussion of feng shui – to the care of succulents.
My other shoulder angles into scepticism. AI’s obsession with bullet points is exasperating, and I’m suspicious that my exclamations and passing conversations translate into adverts on my socials. (Social media company Meta: “It seems like we must be listening to your conversations through your microphone, but we’re not.” [Meta’s emphasis.])
Plus, it can be very non-useful: it once directed me to an air traffic control tower instead of the airport. It’s a dreadful writer, drafting like a 14-year-old pulling an all-nighter for a last-minute assignment.
Yet the pace of AI expansion seems faster than Moore’s law, which predicts a doubling of semiconductor capacity every two years (continually faster, more compact computing). Now, all corporate websites offer visitors the chance to speak with AI. Every AI-controlled switchboard asks callers to speak a few words before redirecting you to the website.
Is AI already out of control and, because we’re so immersed in it, we haven’t noticed?
A bird asks a fish in the ocean, “How’s the water?”
