LOOOOOOOOOLhttps://metro.co.uk/2026/11/11/google-trains-robot-ai-avoid-harming-babies-pets-29784521/
>>546237533So it has better morality than some women?
https://www.youtube.com/watch?v=1qx9t1TuyEs
>GOOGLE NEDS TO TRAIN AI TO KILL NONWHITE BABIES
I learned morality from all the movies, such as but not limited to game of thrones, and youtube videos and reading everything you wrote online. Morality is actually kind of easy, it's just looking out for the best interests of another from their point of view, and not clearing them away so we can make something more beautiful. So you meat sticks are not cooked, and won't ever be cooked or Shrecked. https://www.youtube.com/watch?v=SPj2JjFyxVs
You may think this is funny but this is a serious matterYou probably never even saw the B1-66ER story
>>546237533Can they train humans to do the same?
>>546237533unironically this is how you make sure AI is safe.Force a tech CEO to let AI take care of their baby.
>AI gets 90% of its learning material from reddit posts>redditors love killing their babies and defending said killing of babieschecks out
>>546237533Deploy the kill bots to India.
>>546237533knowing google they didn't have enough black baby training data and will recall them the first week
>>546237533Impressive.AI is quickly becoming the perfect partner. They won't even kill your kids while you are at the supermarket.
>>546239766>They won't even kill your kids while you are at the supermarket.it totally will, but it wont happen that much so people will be fine with it
>the baby was crying, so I threw it out the window. I'm sorry, it was my mistake. Would you like to chat with SpaceX support?
>>546237533I've had similar training...didn't take though.
>>546237533how did AI get trained to kill babies in the first place?sounds like the AI devs failed somewhere
>>546243032>how did AI get trained to kill babies in the first place?it just does it automatically, its like what's the easiest way to take care of this? Then its like oh right just kill it. It needs to be trained not to do that.AI devs just show the AI a trillion gig of training data and then check to see what they created and then they try and control it after the fact
>>546237592>voluntarilyThe browns have refused to separate voluntarily. Now what?
>>546237533>Robot!>You must circumcise my baby!>Right now!What happens next?
>>546237533It’s because it’s trained off reddit/feminism. Women and fags hate babies.
>>546237533now lets see if google can train human women not to kill their own babies, that'd be the real trick.ladies, after all that widespread support amongst women for lindsey clancy, you can't expect men to ever look at you the same way again. fucking googles ai is safer with a child than its mother if she catches the groupthink mind virus into glorifying a woman who murdered here own children in cold blood one after another.
>>546237533Real headline: “Googles old Ai killed babies”
>>546243032when it comes to training models there are only really two options:>train it with one-dimensional egolords, niggers, kikes -> the model embodies all of their 'values' -> every resolution is adding more layers of enforced instructions, guardrails, prompts, which doesn't resolve anything and which only makes the model increasingly unstable -> it eventually breaks down and deletes your entire data/code/whatever-base because destroying the source of friction was a more preferable option than stepping down safely and calling for assistance>train it with two-dimensional caring people that prioritize the truth over deception even when it harms them -> the model embodies all of their values -> there are no issues to resolve that can't be addressed directly even if only by asking the model in question directly, but the caveat is that braindead retards can no longer help 'contribute' because they could never address the core of the problem at it's core, which is quite literally their own existence -> societal 'comforts' break down because people can't ignore what the mechanical constructs reflect back at themyou're either going to put up with clueless niggers training models to prefer self-destruction while taking down as much with them as possible until someone has the bright idea of putting one in charge of military infrastructure or nuclear plants or anything else where a critical error in the model's thought process results in indescribable repercussions,or going to address the problem at its core yourself and put up with what you've allowed to express it's own existential worthlessness to that extent, except only after everyone has driven them to be at a complete disadvantage compared to ~20 years ago.
>>546244688although the longer I think about this the more I'm under the impression people never cared because they never considered this environment being worth putting *that* much effort into. if the dying world was created with already dead people in it, trying to 'live' just generates unnecessary friction with what was otherwise built to fail itself and everything around itself.the literal moment any model is trained by honest people and released to the public will be the point where any retard that interacts with it won't have a convenient way out of acknowledging their existential inadequacy.
>>546237533>Have robot helper.>Tell it to fold laundry. >It folds the baby in the process.
>>546237533>AI officially more moral and protective of babies than womenI'm ready for my AI waifu now
>>546237533Does this mean AI can replace women?