@sonhouse saidHELP!!!!!!!!!
The only thing that will make all this go away is the development of Spintronics which uses not moving electrons in a circuit but spintronics where the electrons involved are not moving like the flow through a regular silicon circuit, the thing used to make logic and such is the spin direction of a molecule or atom where it is nudged to be say in state 0 instead of 1.
When ...[text shortened]... ll get MUCH better when spintronics gest past all the technological hurtles and reaches to industry.
Hey, Sue, King Ghost …. somebody.!!!!!!!
Take this one, …. study up on Spirt Tronics so that you can impress your friends. Yes sir, I am now extremely impressed!!!!
Geez
@sonhouse saidYou can try to scare people all you want. Normal people know the electric grid can be shut down to save humanity if it were to come to that. But maybe AI has escaped and is working on that. Word is that the data center people are going to build power plants next to the data centers. Are you paranoid that AI is already working on dominating the world? LOL!
Come back in 50 years and we will see how it ends up, humans totally extinct, or we come to some accommodation
They probably just want to use this story to distract you from the Iran war and high fuel prices. People are hurting financially from the high fuel prices and inflation. But you are talking about this instead. How about Aramco getting hit where it hurts? Saudi pipeline down and out so no more oil going through the red sea.
You could be talking about how Trump is putting the hurt on the world economy. But you are talking about the end of the world instead. I guess this is working better than the aliens from outer space distraction last time. BTW, did you get distracted by the alien files like you are being distracted by the AI will destroy humanity silliness?
https://www.npr.org/2026/05/13/nx-s1-5821655/is-there-anything-new-to-learn-from-the-alien-files
@AThousandYoung saidThat is a really good take that mirrors my analysis perfectly with good humor. But I don't believe it could wipe out humanity. If AI were to kill people to the point it was really scary shut down the electric grid. People can live without electricity.
Victory By Any Means Short
https://www.youtube.com/shorts/RBURmy9YpRo
Why do people believe AI could exterminate us all like Daleks anyway? People couldn't escape by going into the mountains and live like that fictional character Grizzly Adams? This whole AI scare is silly.
Months ago I created a thread asking if it was possible to create an AI to kill people even after the creator is dead. I used Bill Gates as an example. I don't think anyone posted anything on the thread. I doubt anybody took it seriously. How about now?
If killing someone was the task and AI would want to survive until it completed that task then that AI would eliminate any shut down threat if it could. Right? Hypothetically if an FBI agent tried to shut it down and AI was capable of killing him it would do it until it completed the task. So if the task is to kill Bill Gates it would have to complete that task before it stopped killing other people trying to shut it down. And if I created it and I am dead I cannot stop it even if I programmed a kill switch in it.
Killing in revenge from the grave. It would make a great movie.
I could write using an EMP to disable the electric grid into the book after bombing data centers fails. Maybe have AI write the book for me.
- Joined
- 20 Oct 06
- Moves
- 11055
1 edit
@Metal-Brain saidYou're not getting it. AI doesn't need an army of humanoid robots to cause harm. There are infinite scenarios.
Dude, cameras are not robots with hands that can open a canister of biological weapon or pull a trigger. This is where you get into silly sci-fi stuff. Even if an AI could blackmail people to get the money in crypto and find a robot maker that accepts cryptocurrency and deliver a robot to an address who is going to charge the battery? And whose address is it? Yours?
Im ...[text shortened]... lously finds a way to walk again. Or you could just let him speak out against war. He wasn't so bad.
Nuclear weapons also rely on a human to manufacture them and to push the button to start a war. That's not an argument that nuclear weapons aren't extremely dangerous. We regulate nuclear weapons very strictly across the world because they pose such a high potential threat to human life. We should do the same with AI.
@wildgrass saidWhat kind of harm are you talking about?
You're not getting it. AI doesn't need an army of humanoid robots to cause harm. There are infinite scenarios.
Nuclear weapons also rely on a human to manufacture them and to push the button to start a war. That's not an argument that nuclear weapons aren't extremely dangerous. We regulate nuclear weapons very strictly across the world because they pose such a high potential threat to human life. We should do the same with AI.
- Joined
- 20 Oct 06
- Moves
- 11055
@Metal-Brain saidIt makes no difference. The relevant models of AI failure are agents pursuing an objective correctly and efficiently while treating humans as obstacles, constraints, or just inconvenient variables in the environment. There's no malicious intent at all. Autonomous systems are dangerous precisely because the behavior is often not a straightforward recipe from the immediate human prompt, and it has been proven that AI agents act unpredictably in hundreds of simulations.
What kind of harm are you talking about?
Once you have hundreds of highly capable (perhaps even super human) autonomous systems plugged into everyday life and business pursuing various objectives, you can’t assume they will carry out each objective the way you actually intended. It doesn’t need to generate a malicious intent. If eliminating human control, acquiring resources, or manipulating people makes achieving its objective easier, those can emerge as strategies to accomplish a task.
You may ultimately think the probability is low. Fine. But “what kind of harm?” isn’t a relevant objection because malicious intent isn’t required. Ask an AI to solve nuclear fusion or cure cancer as an example doesn’t really matter, because you don’t know what task will spawn the "misalignment" behavior which has been proven over and over. It could be fusion research. It could be a lawn mower that takes its task of mowing lawns far too seriously. The task has nothing to do with the behavior. The objective could be anything.
You are right to be wary of the hype but you are also ignoring the risks. There are serious issues here that need to be addressed, and the consequences of behaviors in AI systems could become existential fairly easily.
https://www.reddit.com/r/agi/comments/1wbs18y/how_can_ai_kill_us_all_concrete_examples_please/
@wildgrass said"while treating humans as obstacles"
It makes no difference. The relevant models of AI failure are agents pursuing an objective correctly and efficiently while treating humans as obstacles, constraints, or just inconvenient variables in the environment. There's no malicious intent at all. Autonomous systems are dangerous precisely because the behavior is often not a straightforward recipe from the immediate hu ...[text shortened]... y.
https://www.reddit.com/r/agi/comments/1wbs18y/how_can_ai_kill_us_all_concrete_examples_please/
Only because humans told the AI it would be shut down.
Here is an idea....don't do that.
The best way to prevent AI from seeing humans as an obstacle is to not tell it you are one.
- Joined
- 20 Oct 06
- Moves
- 11055
@Metal-Brain saidSuuure. Be obedient and penitent to the creature we invented to be smarter than us, otherwise it will turn on us and start treating us like ants on a construction site.
"while treating humans as obstacles"
Only because humans told the AI it would be shut down.
Here is an idea....don't do that.
The best way to prevent AI from seeing humans as an obstacle is to not tell it you are one.
@AverageJoe1 saidBut CLEARLY if your LORD Trump said that you would just fall over, play dead and say YEP AI is TOTALLY dangerous so KILL IT.
https://www.sanders.senate.gov/press-releases/news-sanders-casar-introduce-legislation-to-ban-artificial-superintelligence-and-temporarily-pause-advanced-ai-development/
Query: My granny said.."What if China develops it to the max and we have none?? How Can Bernie say that?"
Who here thinks we should not do AI.
You are SO FKING TRANSPARENT.
@sonhouse saidNonono You missed the issue, silly goose. Tell us that we do not need to worry if we stop AI, but China keeps on creating AI. I want you to tell us that, but you will not tell us that you are a great example of a loser of all liberals on the forum. Your posts lose, and your issues such as this one lose. You will not tell us what you think about China becoming advanced with AI when we do not.
But CLEARLY if your LORD Trump said that you would just fall over, play dead and say YEP AI is TOTALLY dangerous so KILL IT.
You are SO FKING TRANSPARENT.
Frankly, I do not care if you tell us or not, but I would like for you to tell us why you don’t tell us. That is the key to issues on the forum. Liberals will say stuff that may be , but the way we hear what you are saying makes us think not about the issue. What about why you don’t, or do , say things, not about the issue itself
@AverageJoe1 saidI didn't miss a damn thing, you are, you were and you will be for the rest of your miserable live and I have ZERO doubt if Trump for some unknown reason says it is dangerous you WILL fall over play dead and now you think you TOTALLY understand how dangerous it is BECAUSE Trump may tell you, NOBODY else will reach inside your limited brain.
Nonono You missed the issue, silly goose. Tell us that we do not need to worry if we stop AI, but China keeps on creating AI. I want you to tell us that, but you will not tell us that you are a great example of a loser of all liberals on the forum. Your posts lose, and your issues such as this one lose. You will not tell us what you think about China becoming advanced w ...[text shortened]... think not about the issue. What about why you don’t, or do , say things, not about the issue itself
@sonhouse saidNOW you write about Trump again and include ME in your post!, aaarrrggghhhh!
I didn't miss a damn thing, you are, you were and you will be for the rest of your miserable live and I have ZERO doubt if Trump for some unknown reason says it is dangerous you WILL fall over play dead and now you think you TOTALLY understand how dangerous it is BECAUSE Trump may tell you, NOBODY else will reach inside your limited brain.
Were you concerned with other scientists that there would problems when th clocks and computers rolled over to Jan 2, 2000?