Urgent Plea to AI Firms: Halt the Anthropomorphic Naming of Features, Please!
“I Am Begging AI Companies to Stop Naming Features After Human Processes”
“Doubtless someone out there believes that if we name our AI constructs and processes after human ones, it will help everyone understand what otherwise might seem like complicated technology. But this is a far-reaching assumption, dangerously close to a serious rhetorical touch and go. It’s a faux pas that tech companies have been leaning into a little too hard when they really should not.”
One has to wonder if there’s a secret club of AI companies that meets in candlelit rooms, opting for human-centric terminologies while plotting their next big thing. Do they get a kick out of duping the general public? It’s like replacing car parts with animal names and insisting, “That’s not a spark plug, that’s the woodpecker!”
See, there’s an empathetic tinge in their claims about ‘simplifying’ technology for the layman — painting it as more relatable, less intimidating. Laudable as the intent may be, equating hugely complex algorithms to human functions is, well, kind of absurd. It’s like calling a jet engine a ‘mechanical bird wing’ — sure, it lets you fly, but the similarities end there.
The concept of AI ‘learning’ or ‘understanding’ is a favourite amongst these companies. Yet, when your AI goes rogue and orders 50 pounds of bananas, who do they think will be blamed? It surely wasn’t a conscious choice from your Alexa, despite the misleading terminologies.
And don’t even get started on ‘neural networks’. If artificial neurons mirrored biological ones, neurologists would likely switch careers to become programmers. In reality, plugging some code into an algorithm is hardly on par with the intricate processes occurring in the human brain.
Yes, it is essential to make AI more comprehensive to the non-tech-savvy. But it’s also crucial to not misrepresent the level of understanding these systems have. It creates unnecessary hype and unrealistic expectations, leading to disappointment or worse, mistrust.
It’s only fair to acknowledge the staggering accomplishments in the AI world. But let’s put a pin in misleading nomenclature, shall we? Slamming human attributes onto AI systems does more harm than good. Aim for more accurate terminology that communicates the functions and limitations, without triggering an existential crisis every time Siri misunderstands a command.