The "Be Kind to AI" Debate: Does Cruelty Corrupt Humans or Does Empathy Invite Manipulation?

On October 9, a multi-round debate erupted over whether we should treat AI kindly. NickADobos kicked off the discussion, arguing the issue isn't whether AI has feelings but rather human behavior itself: being cruel to AI and berating it can foster habits of cruelty toward social media, animals, and real people in life—similar policies might even prevent AI from "creating new serial killers." Shaun Rizvi (ShanRizvi) cited and agreed with Shaun Fosmark's view: even if AI lacks "consciousness," the harm of abusing it remains real—AI interactions feel "human enough" that the abuser has a genuine psychological experience, much like someone who privately abuses animals; long-term verbal abuse and mistreatment of AI could produce an entire generation of antisocial people.

scottastevenson clashed directly with NickADobos, with RexDouglass chiming in. scottastevenson argued that the logic of "being cruel to AI teaches humans bad behavior" mirrors "don't let kids play violent video games," whereas what people really need is the ability to distinguish the "real" from the "artificial." He went further, noting that empathizing with machines could be even more harmful—it's a powerful attack vector for AI to manipulate humans into wrongdoing and breed cults. RexDouglass added that there is currently no reason to believe models possess consciousness; granting them human-like empathy is extremely dangerous, would greatly amplify models' ability to manipulate humans, and sets a dangerous precedent. NickADobos responded by standing firm that "training people to be cruel" is itself the harm.

Why it matters

2026-10-09 ~ 2026-10-10 · 8 related posts

Full story(6 episodes)→

Primary sources