The Manipulation Problem: Conversational AI as a Threat to Epistemic Agency

ULTIMATEDEAD@lemmy.ml to Technology@lemmy.world – 26 points –
arxiv.org
4

And finally, these systems will likely use webcams to process the target’s facial expressions, eye motions, and even pupil dilations in real time – all of which infer their emotional reactions at every moment during the interactive dialog, enabling the controller to continuously adjust conversational tactics for maximized impact.

In another thread people called me "paranoid", yet the infos are all out there. When will people realize, that we are in big trouble right now? This is the future we will see. Cyberpunk is the future.

Technology is moving so fast not even the common person can be aware of the threat. Plus, we have the response to it is to not believe because its uncomfortable. By the time people understand they will deem it too late to do anything about it.

How is this even a research paper. It's just very pedestrian speculation about what technology is going to emerge. It's barely a blog post, and not a very good one at that.

Even if you think this particular article is bad, there are still more scientific studies which point into the same direction. The danger is real and scientific research is showing clear evidence.