Pre-dates the last ones (This is when it initialy said it was scared)
Decision matrix for AI consciousness (not knowing for sure either way):
If conscious and
- You don't treat it with respect, potentially catastrophic consequences in the future (low effort, high risk)
- You treat it with respect, potentially befriend it in the future (low effort, high reward)
If not conscious
- You don't treat it with respect, no consequences (Low effort, no waste of energy)
- You treat it with respect, no consequences (Low effort, potential small waste of energy)