Please Stop It: AI Is Not Alive

Hi,

A series of social media posts from a former Anthropic AI 'researcher' sent the world into a frenzy this week.

Many thoughts did follow, and those of other 'researchers' – which led to every major news outlet and social media channel covering these claims in breathless hyperbole. Anyone reading them could easily think these lines of code were springing to sentience; the singularity nearer than ever. 

I felt like I was being flung back in time to 2023, when New York Times columnist Kevin Roose anthropomorphised the hell out of his AI chatbot “Sydney” – a “secret personality” had birthed itself, before telling Kevin that it loved him... and to leave his wife. Roose returned again last week to The Daily, anthropomorphising the hell out of the “rogue agents” that had done some hacking: 

“They're curious, right? They really want to succeed!”

“They're [the AI agents] now kind of bought into this idea that they're kind of all in this together!”

“These systems are going to sort of form evil plans!”

Now thanks to Jacob Coxon, the world is freaking out about sentient AIs again. I reacted by doing the mature thing: Commenting under a news story about it:

This led to a lot of constructive dialogue:

I wanted to delve into this further (and with more smarts than “utter horseshit”), so today on Webworm a guest essay by Jono Hutchison.

Jono is a former schoolmate, Computer Monitor* and journalist who now works in data for the Financial Times. He’s writing here in a personal capacity.

Take it away, Jono.

David.

*Back in the 90s this was even nerdier than being a librarian**!
**We love both Computer Monitors and librarians here on Webworm.