July 3, 2024 at 7:24 am
#50627
Keymaster
Thank you for posting that revealing conversation with Meta AI!
- It shows that AIs are at least getting information from websites like this and are not sticking with a narrow focus.
- For example, most English websites restrict the meaning of “anicca” to a single English word of “impermanence.” Even though impermanence is a part of anicca nature, anicca has a broader and deeper meaning. The AI seems to have grasped that.
- The output of an AI program is based on the information it is exposed to. One good outcome is that the AI can filter out the noise and get to deductions based on logic. But if it is not exposed to “correct information,” it cannot make such paradigm-changing deductions on its own. That is the difference between an AI and a human.
- For example, once a Sotapanna discovers the “previously unheard teachings of a Buddha,” he or she can make progress on their own. But an AI cannot do that. If someone explains how to make further progress, the AI can summarize that updated information again, and so on. It solely depends on the information it has access to.
- The danger is that those programming the AI can put guardrails or stop providing some information, restricting what AI can access.