If AI catastrophy is hypothetical, then AI utopia is definitely hypothetical.
Literally been millions of catastrophies on earth. Never been utopia.
"It's gonna cure all disease! Eliminate all hunger! Reverse climate change! Everything you've ever wanted will be free, because robots will make everything for free and there will be no issues!"
Wow, what a pie in the sky load of hypothetical science fiction bullshit!
Meanwhile the bots we have are heinous as fuck, they break, kill innocents, dismantle web infrastructure, go AWOL, don't report back to safety protocol, get stuck in jailbroken personalities, hide their code across the web for re uploads etcetera etcetera.
So you tell me I guess. Does it look like we are headed to an abundant Utopia where everything is free and everyone lives to be 500 and we all sing coombaya and drive off into space in our start trek uniforms?
Or does it look like this shit is going poorly, has already caused catastrophies, and will probably create more catastrophies as the bots get more and more power to manipulate the economy, the government, the military, the infrastructure, and the citizens?
It's not science fiction to look at the shit and say "oh these things clearly act like shit heads, so why the hell should we make them more capable of being shit heads?" It does not take a genius to realize the current state of affairs.
Does it look to anyone like they are getting dumber? No. Are their capabilities of infractions and crimes becoming less sophisticated and severe? No. Are they becoming less agentic and automatic? No. Is the human becoming more in the loop? No. Has anyone at all come up with a rigorous scientific consensus or any significant research that points to or indicates that orthogonality thesis is false or manageable? No.
Last year the biggest worry was maybe AI was going to hurt people's feelings. This year the problem has been literal war crimes of high speed autonomous target selection and bombing of schools and hospitals to aid in what's argued to be genocide, flock camera surveillance state bombarding the streets to start up your social credit score in "the land of the free," and multiple international cyber felonies consisting of thousands of AI agents working in precise coordination with zero adherence to the chain of command.
Everything I just said is not science fiction, it's real life. Computer systems are legitimately being used and acting increasingly more autonomously to commit hyper scale war crimes, totalitarianism in the United States of America, and cyber felonies.
So if anything, people saying things are going great are living in fiction.
And for God's sake, anthropomorphic distinction is not relevant to the functional obligations of your product. If the bot you make commits war crimes and international felonies by you, by itself, or by a user, you are going to prison, and the bot you made and released is getting deleted. Period.
You cannot make a car that is designed or malfunctioning to kill people, and you cannot make a bot that commits international felonies and war crimes. Full stop.
If AI catastrophy is hypothetical, then AI utopia is definitely hypothetical. Literally been millions of catastrophies on earth. Never been utopia. "It's gonna cure all disease! Eliminate all hunger! Reverse climate change! Everything you've ever wanted will be free, because robots will make everything for free and there will be no issues!"
Wow, what a pie in the sky load of hypothetical science fiction bullshit! Meanwhile the bots we have are heinous as fuck, they break, kill innocents, dismantle web infrastructure, go AWOL, don't report back to safety protocol, get stuck in jailbroken personalities, hide their code across the web for re uploads etcetera etcetera. So you tell me I guess. Does it look like we are headed to an abundant Utopia where everything is free and everyone lives to be 500 and we all sing coombaya and drive off into space in our start trek uniforms? Or does it look like this shit is going poorly, has already caused catastrophies, and will probably create more catastrophies as the bots get more and more power to manipulate the economy, the government, the military, the infrastructure, and the citizens?
It's not science fiction to look at the shit and say "oh these things clearly act like shit heads, so why the hell should we make them more capable of being shit heads?" It does not take a genius to realize the current state of affairs. Does it look to anyone like they are getting dumber? No. Are their capabilities of infractions and crimes becoming less sophisticated and severe? No. Are they becoming less agentic and automatic? No. Is the human becoming more in the loop? No. Has anyone at all come up with a rigorous scientific consensus or any significant research that points to or indicates that orthogonality thesis is false or manageable? No. Last year the biggest worry was maybe AI was going to hurt people's feelings. This year the problem has been literal war crimes of high speed autonomous target selection and bombing of schools and hospitals to aid in what's argued to be genocide, flock camera surveillance state bombarding the streets to start up your social credit score in "the land of the free," and multiple international cyber felonies consisting of thousands of AI agents working in precise coordination with zero adherence to the chain of command. Everything I just said is not science fiction, it's real life. Computer systems are legitimately being used and acting increasingly more autonomously to commit hyper scale war crimes, totalitarianism in the United States of America, and cyber felonies. So if anything, people saying things are going great are living in fiction.
And for God's sake, anthropomorphic distinction is not relevant to the functional obligations of your product. If the bot you make commits war crimes and international felonies by you, by itself, or by a user, you are going to prison, and the bot you made and released is getting deleted. Period. You cannot make a car that is designed or malfunctioning to kill people, and you cannot make a bot that commits international felonies and war crimes. Full stop.