AI security concerns we should all be considering
With the LLM of today still exponentially getting better but still immature in nature the time is now to raise the alarm for what can be done with these. This week Anthropic AI CEO refused the US government on enabling use of there platform for 2 use cases. The first is autonomous weapons, which in itself has some serious consequences.
The second and I think more concerning for all is the use of this technology to enable full scale monitoring of its citizens. Think about the world where anyone get marked or degraded for even questioning the US government. I'm not naive to think we don't have this now, I mean we have terror lists. However to hand all the information to a system which and to use it at will, are we not innocent until proven guilty? Does an LLM model have the right to prove us that way?
I guess only time will tell for how this moves forward but we must urge out representatives and government to move quickly or else we all will live in a place I know I for one don't want to me.
