In one test, conducted by an A.I. safety research group that hooked GPT-4 up to a number of other systems, GPT-4 was able to hire a human TaskRabbit worker to do a simple online task for it — solving a Captcha test — without alerting the person to the fact that it was a robot. The A.I. even lied to the worker about why it needed the Captcha done, concocting a story about a vision impairment. Taken from https://www.nytimes.com/2023/03/15/technology/gpt-4-artificial-intelligence-openai.html

Download link

Notices where this attachment appears

nixCraft 🐧 (nixcraft@mastodon.social)'s status on Thursday, 21-Dec-2023 15:34:24 JST nixCraft 🐧

I have heard that attackers now use artificial intelligence to break the captcha. This explains why every time I change my location, I not only get a captcha challenge but also have to enter random digits from my debit card and push messages for two-factor authentication sent either via the app or via text. This is why we can't have nice things.

In conversation about a year ago from mastodon.social permalink