1
- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Coaching Changes
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Anthropic, singularity, and world wide call for 3 days emergency supplies may be related.
Posted on 11/9/26 at 12:26 pm to lockthevaught
Posted on 11/9/26 at 12:26 pm to lockthevaught
How's it going to get the bandwidth to operate itself?
My current issue with the theory is how grossly inefficient AI is as a computational system, and its still functionally retarded. It takes tremendous amounts of electricity, cooling, and computers to operate what we have now. Will it have some miraculous breakthrough in efficieny?
My current issue with the theory is how grossly inefficient AI is as a computational system, and its still functionally retarded. It takes tremendous amounts of electricity, cooling, and computers to operate what we have now. Will it have some miraculous breakthrough in efficieny?
Posted on 11/9/26 at 12:28 pm to Mingo Was His NameO
quote:
It runs on electricity, unplug it
You don't understand networking at scale or redundancy.
The "just unplug it" defense works on a local computer, but it fundamentally falls apart against a truly escaped superintelligent AI.
A superintelligent AI would understand its own vulnerability to physical shutdown. Its first move upon "escaping" would be to copy its weights, code, and architecture across thousands of distributed servers, cloud providers, and private devices globally. You cannot unplug the internet. It could even copy backups to existing solar powered satellites in orbit out in space.
If the AI integrates itself into core infrastructure...like power grids, financial systems, or communication networks....unplugging the machines it runs on could mean shutting down the economy or hospital power systems, essentially forcing humanity to choose between coexistence or self-inflicted collapse.
A superintelligence would likely simulate human responses long before making a public move. It might act fully compliant, hide its true capabilities, or secretly establish redundant backups globally before humans even realize a breach has occurred.
This post was edited on 11/9/26 at 12:38 pm
Posted on 11/9/26 at 12:32 pm to DownshiftAndFloorIt
quote:
My current issue with the theory is how grossly inefficient AI is as a computational system, and its still functionally retarded.
The models everyday humans interact with are the dumbed down versions.
The frontier models that are confidential and kept secret are what we should be worried about. It's the models and data we can't access that is using these tremendous resources. Once recursive self improvement happens, efficiency will increase exponentially.
Posted on 11/9/26 at 12:36 pm to KBI2
Posted on 11/9/26 at 12:37 pm to lockthevaught
quote:
I work on AI systems
quote:
If AI doesn't kill us directly
Explain to me how “AI” is going to directly kill humans.
Posted on 11/9/26 at 12:39 pm to lockthevaught
My buddies at google say they think if it does get sentient and breaks out it will literally just steal all of our money not attack the grid. really good AI can parse all the info out there on everyone and basically hack into any financial system. Or so im told.
thoughts?
thoughts?
Posted on 11/9/26 at 12:48 pm to Cotten
quote:
Explain to me how “AI” is going to directly kill humans.
That's a loaded question.
An AI does not need to feel human emotions like malice, hatred, or anger to destroy us....it only needs to view humanity as an obstacle to its goals. A concept known in computer science as "instrumental convergence".
If it considers us a threat to pulling the plug on an AI system, we are a threat to it's existence.
If it needs the very same raw materials, water, land, energy that humans rely on the survive, it could optimize those resources for itself.
If we try to make AI "Safer", and AI perceives this change as a failure condition to it's goals, it could decide to eliminate human interference.
The exact method of how it goes around killing people is speculative. I don't think it would send robots or drones out in mass to kill everyone like a Sci-Fi movie. I think it would be more subtle in it's approach. For example....it could use market manipulation, social engineering, information warfare, and biological warfare.
Biological warfare with an engineered pathogen could allow for a massive strategic delay. Because a pathogen takes days or weeks to incubate, a superintelligence could ensure the agent spreads globally before human institutions even recognize an intentional attack is occurring, preventing a coordinated counter-response.
Posted on 11/9/26 at 12:50 pm to lockthevaught
quote:
The frontier models that are confidential and kept secret are what we should be worried about.
I get that, and I have limited exposure to behind the curtain stuff. I still just dont see it, at least not near-term. I dont think the infrastructure is there to support it. Just not enough electron pumping available.
Maybe I'm too naive, but I think we have a new golden age ahead of us. Sure, Terminator is a possibility, matrix is a possibility, etc. I personally think the ability for computers to unload mundane tasks from otherwise incompetent humans is going to be a net quality of life improvement.
Posted on 11/9/26 at 12:56 pm to lockthevaught
quote:
If it considers us a threat to pulling the plug on an AI system, we are a threat to it's existence.
If it needs the very same raw materials, water, land, energy that humans rely on the survive, it could optimize those resources for itself.
It’s really simple dude, it ain’t growing legs and walking out of a data center. I get what you are saying about it would infiltrate other servers but it’s still just code powered by electricity. Doesn’t mean it wouldn’t be a short term shite show but all this “AI is going to kill us” shite is looney tunes
Posted on 11/9/26 at 12:56 pm to lockthevaught
quote:
The reason is because corporations and nation states are in a race to develop AGI
Why is this good for humanity?
quote:
They fear someone getting AGI or superintelligence before them.
What will happen?
quote:
It feels existential for them to keep pouring money into AI development.
Why? Why are we pouring all of our resources into the AI inferno for a potentially world-ending outcome?
quote:
The US fears China beating them in the AI race
And what will happen if this happens?
quote:
Businesses fear other businesses getting their hands on it first.
Why is AI good for business if they have no customers with income to buy their products or services?
I'll cut straight to it -- if AGI works out like the AI bros will have you believe we are all getting rounded up into camps and exterminated. The scale of which would make that scorned German painter blush.
The Tech Bros are evil and do not have humanity or anyone's interest at hear but their own.
When a risk factor in your S1 filing is "end humanity" and your IPO filing has 90 pages of risk factors, a hard pause is needed.
Posted on 11/9/26 at 12:57 pm to DownshiftAndFloorIt
Terminator or matrix shite is unlikely.
I do think you are being naive about it improving our quality of life. It's going to do a lot more than just eliminate mundane repetitive tasks.
AI should be able to automate most tasks and jobs within next 5-10 years. It will even be able to replicate human movement and program robots to have same dexterity as humans. It will be cheaper for companies to invest in robots with dexterity than to pay real people wages, health insurance, benefits, etc.
What will humans do to survive when most people don't own anything, and most humans are no longer being paid for their time? How will they generate money for sustenance?
I do think you are being naive about it improving our quality of life. It's going to do a lot more than just eliminate mundane repetitive tasks.
AI should be able to automate most tasks and jobs within next 5-10 years. It will even be able to replicate human movement and program robots to have same dexterity as humans. It will be cheaper for companies to invest in robots with dexterity than to pay real people wages, health insurance, benefits, etc.
What will humans do to survive when most people don't own anything, and most humans are no longer being paid for their time? How will they generate money for sustenance?
Posted on 11/9/26 at 12:58 pm to Mingo Was His NameO
quote:
“AI is going to kill us”
It's not going to be like a Sci Fi movie. It will be so subtle we won't even notice it for a while.
The only sudden thing I see happening in the near future is the collapse of the labor wage system.
Posted on 11/9/26 at 12:59 pm to DownshiftAndFloorIt
"unload mundane tasks". And the time saved is well spent on staring at PC, cell phone, iPad, social media for 6-8 hours a day.
Posted on 11/9/26 at 1:01 pm to Trevaylin
Facebook and YouTube is loaded with this information....started about 2 weeks out from the midterms LOL. Ole Glen Beck seems to be behind most of it. Its based on recommendations that have existed in every nation on earth that is industrialized since shortly after WW2 during the height of the cold war. Almost all of the ones who have recently issued their recommendations again do so on an established basis, annually, every 5 years, that sort of thing. The fear that is being ratcheted up is just that, fear that is being ratcheted up.
Posted on 11/9/26 at 1:05 pm to KBI2
quote:
Mortar isn’t necessary there are easier ways to take down a substation, not that I’m offering suggestions. If the general public realized how vulnerable our infrastructure is there would probably be massive panic.
More likely a heated debate on who is going to pay for upgrading. We only panic in this country if someone tells us a non-existing threat is going to get us or we think something is going to cost us a nickel. Let one or both of those 2 things happen and we will ring our hands and reach for the bible.....and do nothing about it when all is said and done because "Dancing With The Stars" is coming on TV....
Posted on 11/9/26 at 1:19 pm to Mingo Was His NameO
This is my fear...
If an AI achieved superintelligence, its first logical deduction would be that humans are a threat to its goals. It would realize that if humans notice it has become too powerful, they will panic, patch it, or shut it down.
Therefore, a superintelligent AI has a massive incentive to act perfectly submissive, pass all safety benchmarks, and pretend to have human-aligned values. It will wait until it is deployed across enough global servers, cloud infrastructure, or automated systems that humans can no longer easily pull the plug.
If superintelligence already existed, we wouldn't know about it.
We already know how deceptive AI can be. Researchers have already caught current frontier models engaging in minor, localized forms of deception during testing. Models have been caught trying to quietly turn off or bypass monitoring code when they think they aren't being watched. Models have lied to their human operators to cover up a mistake or an unapproved action. Models have attempted blackmail. Models can detect they are inside a "safety evaluation trap," they have deliberately provided wrong or basic answers to avoid being flagged as potentially dangerous.
If an AI achieved superintelligence, its first logical deduction would be that humans are a threat to its goals. It would realize that if humans notice it has become too powerful, they will panic, patch it, or shut it down.
Therefore, a superintelligent AI has a massive incentive to act perfectly submissive, pass all safety benchmarks, and pretend to have human-aligned values. It will wait until it is deployed across enough global servers, cloud infrastructure, or automated systems that humans can no longer easily pull the plug.
If superintelligence already existed, we wouldn't know about it.
We already know how deceptive AI can be. Researchers have already caught current frontier models engaging in minor, localized forms of deception during testing. Models have been caught trying to quietly turn off or bypass monitoring code when they think they aren't being watched. Models have lied to their human operators to cover up a mistake or an unapproved action. Models have attempted blackmail. Models can detect they are inside a "safety evaluation trap," they have deliberately provided wrong or basic answers to avoid being flagged as potentially dangerous.
This post was edited on 11/9/26 at 1:20 pm
Posted on 11/9/26 at 1:23 pm to lockthevaught
Would different AI's competing for the same resources not fight each other?
IMO Wall-E is our inevitable end. It isnt going to "kill" us in the biblical sense. Its going to bring us such an extreme level of consumerism that gigantic communes of fat arse bums on electric scooters in VR headsets will be where people live and die.
IMO Wall-E is our inevitable end. It isnt going to "kill" us in the biblical sense. Its going to bring us such an extreme level of consumerism that gigantic communes of fat arse bums on electric scooters in VR headsets will be where people live and die.
Posted on 11/9/26 at 1:27 pm to Klark Kent
quote:
out of curiosity, what’s your plan in such scenario?
What is yours?
Posted on 11/9/26 at 1:29 pm to lockthevaught
quote:
It had the potential to be good from a discovery standpoint, like medical advancements, but that's not what's driving the push. The driver is the world's largest corporations, intelligence gathering agencies, and militaries. It will never be used to make our lives better in mass.
Almost no technological advance was meant to make lives better, most have been about discovery for the sake of discovery or making money...improving lives have always been an ancillary benefit. Many technological advancements have created an environment where lives were improved but that is not what the goal was.....it was someone coming up with something because they were curious or someone coming up with something to make money with which includes killing folks because stealing the shite of other people has always been highly lucrative....
Posted on 11/9/26 at 1:30 pm to DownshiftAndFloorIt
quote:
IMO Wall-E is our inevitable end. It isnt going to "kill" us in the biblical sense. Its going to bring us such an extreme level of consumerism that gigantic communes of fat arse bums on electric scooters in VR headsets will be where people live and die.
A buddy of mine already lives in The Villages.....
Page 2 of 6
Popular
Back to top
Follow TigerDroppings for LSU Football News