We should fear AI

Sci Fight: Science/Comedy Debates

Sci Fight is a quarterly Science Comedy Debate where scientists and comedians come together to debate serious issues in a ridiculous manner.

This round's topic: We should fear AI.

AI is all up in your business. It knows when you’ve been sleeping. It knows when you’re awake. It knows things about you that would bring a family bingo night to an awkward halt. And there’s new ways it's being used every day; AI may soon be used in judicial sentencing, in medical diagnoses and to care for the elderly. AI is able to be programmed out of the subjectivities and biases of us fallible humans, and it eats far less crumpets.

Some things about AI are simply annoying; like the pro-Russia twitter bots liking your posts (and never buying tickets to your shows) and the weird social media algorithms that decide to start advertising you menopause remedies or combine harvester repair kits. But then there are the scarier moments; when you’re reading or looking at something that should be made by a human, but it just doesn’t have the right number of fingers. ‘Something seems off’ doesn’t feel like an enduring defence against AI’s ever increasing sophistication; Uncanny Valley was three stops ago.

Academics downplay ChapGPT’s efficacy; as when you ask it a question it spontaneously hobbles together a series of roughly correlated facts it found elsewhere to make a hollow semblance of an argument and that's definitely not what I spent my undergrad arts tutorials doing. Can schools and universities genuinely invigilate AI out of the classroom?

While humans immediately start making those throaty noises about AI coming for our jobs, the same way they did with the invention of the washing machine, the street sweeper and the bidet (okay, less that last one) AI may pose threats to humanity that we’re not even across yet. Will the singularity keep us up to speed what it's up to? Should we be keeping our friends close and AI closer?

Or should we simply be celebrating AI? If all AI is is the enhanced and refined networking from the vast product of human output- then maybe AI is more human than humans! If that’s the case, then is all we really have to fear is…. ourselves?