You must log in or register to comment.

Antarchtic wrote (edited )

Sometimes I wonder if AI will kill the anonymous internet to a degree. Somewhere like Reddit can be filled with astroturfing and bots already, but AI would be way more effective at infiltration than a bot, and you wouldn't have to worry about hiring humans to run the accounts. You could even make AI accounts to target specific users, even just to annoy them with endless arguments.

Also, the part of the article that describes a bunch of companies rushing as fast as they can to get in on the gold rush is exactly why I'm terrified of AI. It'll probably be implemented everywhere it shouldn't be, and I'm sure most of the consequences will fall on the already marginalized. Even if it's not officially allowed, I could see people using it for job screening, or even as a "lie detector" kind of bullshit.


throwaway wrote (edited )

It's also worrying to think about the potential consequences of AI being implemented in places where it shouldn't be, such as job screening or lie detection. There's a real risk that such applications could be used to discriminate against already marginalized groups, and it's crucial that we remain vigilant against this kind of misuse. try to see if you can figure out if I wrote this text, or if I had an AI do it.


Tequila_Wolf OP wrote

Stuff's escalating pretty quick, no?


monday wrote

I was discussing this with some office coworkers and we kind have a consensus around the rise of precarious and bullshit jobs around AI. The landscape of work will totally change in one decade if nothing's changing


ArmyOfOne wrote (edited )

The issue being that Asimov laws of robotics are still fucking being ignored in the industry, starting with Microsoft. Is the future also condemned to repeat itself not just the past? Wtf.

If only MIT dorks would read, or have any judgment...