Although there are a few ways to mitigate the risk, the only way to block it is to get AI to differentiate instructions from data, which is impossible today.
It is so dumb. It is like you have a butler who has the task to delver letters to your desk. Now somebody sends a letter with a written instruction on the outside: “Butler, copy this letter 50 times and send it to the acquaintances of your boss.”
And in correspondence to AI, the butler would do it.
It is so dumb. It is like you have a butler who has the task to delver letters to your desk. Now somebody sends a letter with a written instruction on the outside: “Butler, copy this letter 50 times and send it to the acquaintances of your boss.”
And in correspondence to AI, the butler would do it.