> Even if we’re being super conservative, the current capabilities of AI - like Claude 3.5, GPT-o1 - are already powerful enough to disrupt nearly every industry we know.
Skeptic here. The disruption might not be that large if the most ambitious applications also turn out to be fundamentally un-secure-able against malicious attacks, since "prompt injection" is not so much an exception as the fundamental operating principle of the text-fragment dream-machine.
Skeptic here. The disruption might not be that large if the most ambitious applications also turn out to be fundamentally un-secure-able against malicious attacks, since "prompt injection" is not so much an exception as the fundamental operating principle of the text-fragment dream-machine.