3 ms·
>I think we'll be seeing AI-driven scammers, calling and messaging us, almost incessantly, I think we'll be seeing AI-generated political manipulation, like fak
by asdff 4y ago
>I think we'll be seeing AI-driven scammers, calling and messaging us, almost incessantly, I think we'll be seeing AI-generated political manipulation, like fake videos and audio, of politicians and public figures, I think we'll be seeing AI-driven promotional campaigns, and almost undetectably subtle media manipulation, like corporations seeding news stories with their talking points. I also think that we'll be seeing AI-driven workplaces, where it will be hard to be simply human, and AI-driven surveillance, by authoritarian regimes (and people right here, will happily sell it to them).
I've been having similar concerns, however, I just wanted to note in this paragraph there is not a damn thing you've listed here that isn't already being done right now without ai. We already get scammed incessantly. Every political message we receive is meant for manipulation versus actual understanding. Corporations have been seeding news stories since there was ever news stories. We have corporate cultures that routinely dehumanize us, forcing us to be sedentary indoors, when we ought to be foraging during the daylight if we actually lived according to our biology. we surveille eachother, and have no need of any advanced technology to oppress ourselves by the millions with brutal effectiveness. If you are concerned now, then you must not have been paying much attention. If you have issues with these things, then ai is not what you need to be going after, this is how the world is incentivized to run.
- ChrisMarshallNY 4y agoMany of these scams are older than dirt. I remember, in the 1980s, companies getting actual snail mail 419 scams, then, in the 1990s, faxes. But AI will be giving the same sharks, new tools. Soon, you will actually be hearing your granddaughter on the phone, screaming, as they rip out her fingernails. The "IRS agent" will no longer have a dodgy Indian accent. Those "I caught you masturbating" emails will now have actual deepfakes of your face. In fact, I wouldn't be surprised if a new racket appears, of extortion, using wholly-fabricated nude/embarrassing photos. I suspect that influencers will be targeted, as they put out so much material, and many have means. I just know exactly how bad people can get; better than most, hereabouts. I have few illusions. But we'll also be getting some good. This is the real deal. It isn't Claptrap's BECHOWafers.
- asdff 4y agoIf the scams become widespread people will react accordingly. Real kidnappers ripping fingernails for ransom will find that they get hung up on and now have to resort to other strategies to get past the spam and get payment. People will realize how easy it is to make a deepfake and just not care about photos of them in compromising situations since they could now always claim plausible deniability. Gullible people will fall victim of course, but it doesn't take much even today to have someone hand over their life to you with some social engineering. There's probably millions of bots that were doing just that with some canned responses and getting a good enough return to continue hosting the bot farm at least even before chatgpt. I doubt the internet scammer needs much sophistication to get a decent return.
- yencabulator 4y agoIt's a question of scale. Back in the day, a three-letter agency could send a person to spy on your 24/7. They couldn't do that for everyone. Now, mass surveillance is possible. Back in the day, you had to send a letter in the mail or have a person calling someone to run a scam operation, and that had a cost to it. Email started changing that, but remained impersonal, non-interactive. Generative ML changes that. Back in the day, traveling across the US took months or even years. Now, you can fly across in a few hours, or drive across in a few days. You're behaving as if that didn't change things.