Google's statement that it includes a watermark that can identify them as AI-generated is completely braindead as to how disinformation works. You think people spending literally three seconds to derail a discussion give a shit about that? This is so easy to abuse www.404media.co/google-earth...
Публикация
Took me seconds to make an image showing 'protesters' around Google buildings, using Google Earth's new AI generation tool. You can make bomb blasts, protesters, drone strikes, nuclear plants. How on earth did Google think this was a good idea www.404media.co/google-earth...
Обсуждение
32 прямых ответов · 142 сообщений
That was fast. Google sent me a statement saying it is now “rolling back“ this AI feature after our demonstration and others from OSINT researchers www.404media.co/google-earth...
Google Earth’s New AI Lets Anyone Fabricate Completely Bullshit Satellite Images"Tonight I typed just one sentence into Google Earth and put refugees near the Mexican border. Then I planted a nuclear plant in Iran. Then I put a fatal crash on a street in Amsterdam."www.404media.coMy theory: bsky.app/profile/munt... They attitude is that the risk is not in releasing a harmful AI feature, the risk is in *not* releasing enough AI features.
munteanuaThe incentive, if you work at one of these tech companies is to go balls deep on AI. If you add AI features to your product you will get promoted, you will get recognized you will get paid more. On the flip side if you don't do enough AI you get canned so the company can buy more GPUs.
Yeah, my original career was teaching adolescents w/behavior disorders. Now I teach teachers to do that. I had one call me in tears b/c a kid stood on his desk & mooned her. She was mad I never told her they might do that. 🤷♀️ I can’t provide a list of everything they MIGHT do. If it’s possible, yep.
I don’t see a watermark either! Some of the debris is weird tho, like too many rounded wavy edges and that would be the only thing that would tip me off to question its validity… if I even looked that closely before I reposted it! 😬 But G👀gle’s fixing that in the AI right now as it reads my post… 😮💨
1.Iran - more likely to be fact checked but open to abuse at highest level 2.Migrants - swayed to far right, deemed as less users abusing overall to meet an agenda 3.Amsterdam - signals anyone, anywhere for any agenda could abuse reality & create a believable situation- to me this was the scariest
D
Shut. It.Down. @booker.senate.gov @corybooker.com @andykimnj.bsky.social
Oh, I think we know how Google's people thought it was a good idea. They simply didn't think at all
Why not just hand out the shovels at this point? Dark times.