@cendyne That's what made me anxious the first time I got to play with GPT-3, I knew it was big shit and it would become hard to tell humans apart from bots.
Now I occasionally click on a video with an obvious AI voice, with text generated by ChatGPT, or find a comment section that looks very sus. This is still very primitive stuff done by lazy spammers, it's only a matter of time until someone creates more sophisticated tools that are fully automated.
And we're just getting started, these models are gaining vision, and there's no real barrier to giving them more modalities to become even more natural without relying on TTS.
Making CAPTCHAs is becoming more difficult.
And then the CEO of "Open"AI comes out with a crypto grift that involves the scanning of people's irises to "address" that in a backwards way.
Now just wait until these advances get used for killing by military and police somewhere and we've got ourselves a complete dystopia, and it's not even cool like in cyberpunk, it's mostly spam.
@cendyne We've been using automation-based tools for years to spam people (robocalls, emails, etc.) so it makes perfect sense that AI would be harnessed to do the same thing. Legality, ethics, and common decency be damned when there's a possibility of money to be made. :blobfoxdead:
@RoboticistDuck@cendyne@cetsch I’m just thinking of the scenes from the “Terminator” franchise where the cyborgs are talking on the phone with a clearly fake voice. (“I love you too, sweetie” and “What’s wrong with Wolfie?”)
If I ever get a wierd call I’d just say something nonsensical and see how the family member replies. We also have a “under duress statement” (phrase that sounds plausible but totally false for our family) in a serious emergency.
Add comment