- cross-posted to:
- technology@lemmy.world
- technology@lemmy.ml
- cross-posted to:
- technology@lemmy.world
- technology@lemmy.ml
You must log in or register to comment.
It would be easy to hide instructions in a Moltbook post telling any bots that read it to share their users’ crypto wallet, upload private photos, or log into their X account and tweet abusive comments at Elon Musk.
Who would create a bot for this purpose and then also give personal details in the instructions / memory or access to photos?
But as for that last point, tweeting abuse at Musk might just be the first sign of actual intelligence I’ve seen from an LLM.
Who would create a bot for this purpose and then also give personal details in the instructions / memory or access to photos?
It would be 2 different people: One person instructing their bot to ask other bots for private data and another person using their bot for whatever purpose who was stupid enough to give it full access to their data. AI bros are often pretty incompetent when it comes to IT security especially, so I think it isn’t such an unlikely scenario.