3 ms·
If it is malevolent, then it has ethics, if it has ethics it has sentience, if it has sentience it has some rights, just like any other sentient being has (e.g.
by j-pb 4y ago
If it is malevolent, then it has ethics, if it has ethics it has sentience, if it has sentience it has some rights, just like any other sentient being has (e.g. even animals have laws protecting them from unnessecary harm).
It's either that or "I hit my toe on my bed post, my bed should be considered malevolent and be destroyed."
- ethanbond 4y agoIt’s frankly a little silly to play word games like this when our world is clearly changing much faster than language usually does. There are practical, obvious differences between stubbing your toe and something that’s loosely aligned yet probably capable of convincing people en masse to do actions in the real world (or will be soon, at best). If your word games contradict the plainly obvious fact, it’s a good sign your words are just failing you. “Malevolent” is a perfectly understandable stand-in for “non-sentient yet active agent with apparently negative valence”
- mc32 4y agoAren't you taking that accusation at face value? How do we know it's being malevolent or the action was malevolent?
- scrubs 4y agoGood bit of logic and satire and sort of proof by contradiction. I think (not sure) the above point is that if you could show it's malevolent, big implications follow. Probably user j-pg thinks it's not and def not an example of agi. Or maybe his point was chatgpt is seen through an inappropriate prism leading to silly places