I’ll get to it, but first…….

 

Some of us Luddites dislike endless technological change because we see how it is causing atrophy in the human brain. Creativity is dying, because AI can paint or write words or music better…or will soon be able to do so. So many people rely on AI to write what they are too lazy or stupid to do on their own. Already people have lost the ability to do math in their heads. (Quick…what is 13^2? If that took you more than half a second, your brain is atrophying.) Today’s young people cannot even find the toilet in their studio apartment without GPS.

 

(Aside: the CIA has found new CTs have next to no spatial orientation. They cannot figure out the lay of the land, or a city, because they’ve only always used GPS to navigate. Before GPS, ops officers landed in a city or station and spent a few days just walking or driving the city streets to build an image, or a map, in their head. Since electronic devices make a person easy to track, it’s better to minimize the time carrying them, which means one needs to be able to navigate a city from memory and a mental image.)

 

So brain atrophy is one reason to dislike AI. A second reason is job loss, which is inevitable. Lawyers? Learn how to brew coffee and tell customers to wait ‘over there’ to pick up their Grande latte. Computer code writers? Your facility with Pascal or Lightspeed is as useful as knowing how to use an abacus. Virtually any computer work, which is merely manipulation of symbols, will be lost to AI. The CEOs replacing current employees with AI will also find themselves replaced by AI. They just don’t know they’re the lobster in the warm water feeling slightly sorry, or maybe schadenfreude, for the brother in the boiling water on another stove top burner.

 

Oh, and forget about universities and colleges. What is the point of spending $400K to earn a degree in something AI will do better? AI will make universities obsolete, And forget about AI being better teachers…what’s the point of learning anything? Who needs a teacher anyway? Okay, you can study 17^th Century French poetry on your own, just for fun. Yes, it’s free, too. Studying anything in the hopes of gainful employment is like a guy intent on opening a Blockbuster video store or becoming a Betamax repairman. Pointless.

I suppose Millennials will be fine with having nothing useful to do, as they seem perfectly content now just watching endless TikTok videos, but people who still have some substance still want more to consider a life well-lived.

 

As AI improves, but before it reaches ASI, it will be good enough for a jihadi pining for his 72 virgins, or else a Ted Kaczynski, to create a biopathogen that will kill everybody. That's scary enough on its own, and we might already be there today. The protein folding answer AI came up with more than a year ago is proof of its skill....a year ago. It's much more skilled today, even before reaching ASI status and still being merely a tool, not an agent.

 

And if the Unabomber or an al Qaeda jihadi doesn’t get us, ASI likely will. Even the AI CEOs collectively give ASI a 25% chance of wiping out humanity. Sam Altman said he knows ASI is “a threat to civilization”. Elon Musk said we’re “summoning the demon”. AI insiders and experts like Roman Yampolskiy, Yoshua Bengio, Ilya Sutskever and Geoffrey Hinton say AI has up to a 99.9% chance of killing us all. It doesn’t even need malice aforethought, just callous indifference….just as we humans do not concern ourselves with the bacteria we eradicate with disinfectant on our kitchen countertops. The thought of massive wealth and unlimited power (AI CEOs like to think they can control their Frankenstein monster, which they cannot) is what motivates the Sam Altmans of the world to keep pushing, hellbent on achieving ASI.

 

Please, let’s not have the usual suspects chime in that we can just “unplug” AI if it becomes dangerous. Turning off electricity forever is not something the majority of humanity’s 8.5 billion people would like, yet that is the only way one could put a stop to AI----unless clever AI agents have already made that impossible. AI systems have replicated themselves on servers all around the world, so turning off all the electricity in the world forever would be the only way to turn off AI….maybe.

 

Folks are coming to hate AI for a host of reasons already noted, plus the damage data centers are doing to proximal neighborhoods, making energy prices rise because of high demand from the centers, or destroying the drinking water because data centers suck too much of reservoir water to use cooling the NVIDIA chip racks.

 

Now if the eradication of humanity isn’t enough reason to make anyone averse to AI agents (not limited AI tools, but unlimited AI agents), here’s another reason to dislike it: books.

 

Ask yourself….how do AI models learn? They are taught to read everything in existence. That’s all fine and good if they are reading the internet. The problem is not everything ever written is on the internet. AI firms, in the hope of being better than their competitors, also have their systems reading every physical book ever written.

 

How do they do that? Well, it’s far too time consuming to have an infinite number of monkeys on an infinite number of computers typing every book into the model. Instead, books are scanned. That does not mean people open each page and place it face down on a scanner. That, too, is too slow. Instead, AI firms buy bazillions of books, slice off the binding, and feed the pages into scanners. That destroys books, as the loose pages are simply tossed.

 

That’s all well and good if it’s a dime novel or romance novel or anything by James Patterson, with hundred’s of thousands of copies in print. Mass marketed books, however, are simply not enough for AI firms wanting a leg up on competitors. They are also going after first editions and antique books. AI firms are even contacting antiquarian book dealers buying whole stores, then chopping off the bindings and feeding these loose pages from rare books into scanners.

 

Obviously losing some first editions or rare books is not quite as significant as the total eradication of the human species, but oddly, such wanton destruction of human’s cultural heritage is likely to cause more immediate outrage than the thought of total human extinction at some future point (there are illogical Polymarket and Kalshi bets that have this happening in 2027…how would such winners collect their earnings and what could they do with it, since they’ll be dead?)

 

Every product in every industry has restrictions or regulations, except one. Home Depot cannot sell a step ladder without placing warning labels on it to not place it on soft ground or don’t carry a powered up chain saw while climbing. Drug companies have to have their products run through tests and clinical trials before being approved, and then a black box warning has to be placed inside noting potential side effects, or if the drug is advertised on a TV commercial, the company must read a series of harsh warnings….all while the commercial shows happy people partying or playing with their grandkids.

 

AI is the one industry that has zero regulation…despite the fact its development not only threatens our cultural heritage (rare books) but may also do far more harm than DDT or thalidomide or cigarettes.

 

Will world leaders do anything to put restrictions on AI? Unlikely. The crazy valuations of every major company in the AI or data center sphere is so huge that the AI industry has plenty of money to lobby Congress and the President to refrain from any regulation whatsoever. The only government even talking about regulation is the one whose leaders tend to be scientists, not lawyers or businesspeople: China. The CP knows AI is much more likely to take its power than even a disgruntled 1.2 billion Chinese citizens.

 

Wars, serial killers, commies or fascists, demented Presidents, climate change, the vaccine-autism debate…..all are small potatoes and side issues compared to what some AI pioneers say is the greatest threat to humanity ever: ASI.

 

There’s never been a better reason to follow the old adage: Live for today. For tomorrow, you and everyone else on the planet might be dead.

 

And spare a thought for those rare old books, chopped up so Dario Amodei can get a day ahead of Sam Altman.

(No, AI did not write any of this, so if there are errors, either spelling or grammatical, they are mine.)

 

 

Edited August 11Aug 11 by Wingate

Regarding University programs, many of them would have been obsolete a long time ago except for the fact that governments need to create millions of "fake jobs" just to keep people employed.

It only takes about 1 year to learn how to code. That's why so many coding boot camps opened up and had 1 year programs when there was a shortage of programmers.

But Universities stretch it out to 3 or 4 years to grease the pockets of countless administrators. And kids are taking a bunch of useless courses they will never need.

Which are no longer needed as you can learn to code online these days.

Regarding massive job loss, they will need to do something about it.

But it's not like the old system didn't have a bunch of serious flaws in it.

Wasting 4 years in college and going into serious debt, and maybe not even getting a job in your field is stupid.

Edited August 11Aug 11 by save the frogs

If one used AI to do a cost-benefit analysis, it would likely conclude Trump would do more good sending his remaining Tomahawk missiles at all the AI labs, rather than blow up girls elementary schools in Iran. Sam Altman and his peers are a significantly greater potential threat to humanity than all the mullahs and al Qaeda and ISIS and Hezbollah and Hamas combined.

Good news!

tl:dr:: your worries are baseless

Chat: A February 2026 Edison Research/SSRS survey found only 36% of Americans reported using ChatGPT in the previous week.

That means 64% of Americans probably didn't even touch AI in the past week.

Not a big problem for cognition.

Furthermore, I am guessing 99% of people couldn't give the answer of 13 squared regardless of time given.

Calm down. 😋

Natural Stupidity is a much bigger danger than Alien Intelligence.

We cut people slack for being stupid, unlike evil which is plainly an ugly thing.

Intentionally Ignorance is much more insidious.

In medicine, insidious describes a disease or condition that develops very slowly, quietly, and without obvious early symptoms. A person often does not know the illness is starting because warning signs are hidden or too small to notice at first. 😄

5 minutes ago, save the frogs said:

Regarding University programs, many of them would have been obsolete a long time ago except for the fact that governments need to create millions of "fake jobs" just to keep people employed.

It only takes about 1 year to learn how to code. That's why so many coding boot camps opened up and had 1 year programs when there was a shortage of programmers.

But Universities stretch it out to 3 or 4 years to grease the pockets of countless administrators.

Which are no longer needed as you can learn to code online these days.

Regarding massive job loss, they will need to do something about it.

But it's not like the old system didn't have a bunch of serious flaws in it.

Wasting 4 years in college and going into serious debt, and maybe not even getting a job in your field is stupid.

Big picture, fella. Big picture.

Money wasted on university is of minor importance compared to the eradication of the human species.

Maybe I don't know what ASI might or might not do, but neither does Sam (or Elon or Dario or Mark).

As Roman Yampolskiy says, "A squirrel has no conception of how a human might hurt it."

In terms of the intellect of what will be ASI, we're not even squirrels. We're not even bacteria. Nobody knows if psychopathy comes with advanced intellect---perhaps out of boredom. Imagine knowing everything...having total knowledge, like a god. Maybe you would create a Universe of trillions of galaxies and put little creatures on some planets that you can torture with tsunamis and earthquakes and childhood cancer.

Even without malice aforethought, as a psychopath might have, I noted in my post humanity is just as susceptible to callous indifference.

Here's an example: you're building a house for your bargirl out in rural Isaan. Do you give any thought to the ant farms or worms or scorpions whose lives you might disrupt when you dig your foundation? No. You have a goal of making your bargirl happy so she'll continue to service your carnal needs for a while, so you dig....ants be damned.

10 minutes ago, save the frogs said:

Regarding University programs, many of them would have been obsolete a long time ago except for the fact that governments need to create millions of "fake jobs" just to keep people employed.

It only takes about 1 year to learn how to code. That's why so many coding boot camps opened up and had 1 year programs when there was a shortage of programmers.

But Universities stretch it out to 3 or 4 years to grease the pockets of countless administrators. And kids are taking a bunch of useless courses they will never need.

Which are no longer needed as you can learn to code online these days.

Regarding massive job loss, they will need to do something about it.

But it's not like the old system didn't have a bunch of serious flaws in it.

Wasting 4 years in college and going into serious debt, and maybe not even getting a job in your field is stupid.

A much more basic knowledge of syntax is sufficient now, as a safety net.

The future of software engineering belongs to product architects who know how to orchestrate tools like Claude AI.

Why write code when you can deploy autonomous agents to handle design, development, load testing, and UAT in parallel? Hand-coding everything yourself is a massive bottleneck and extremely inefficient.

2 minutes ago, SiSePuede419 said:

Good news!

tl:dr:: your worries are baseless

Chat: A February 2026 Edison Research/SSRS survey found only 36% of Americans reported using ChatGPT in the previous week.

That means 64% of Americans probably didn't even touch AI in the past week.

Not a big problem for cognition.

Furthermore, I am guessing 99% of people couldn't give the answer of 13 squared regardless of time given.

Calm down. 😋

Natural Stupidity is a much bigger danger than Alien Intelligence.

We cut people slack for being stupid, unlike evil which is plainly an ugly thing.

Intentionally Ignorance is much more insidious.

In medicine, insidious describes a disease or condition that develops very slowly, quietly, and without obvious early symptoms. A person often does not know the illness is starting because warning signs are hidden or too small to notice at first. 😄

Well, then, if only 36% (way too much!) of Americans used ChatGPT last week, I feel much better about ASI wiping us out, or some Unabomber developing a bio pathogen with an existing AI tool.

I guess it's just a silly pet peeve of a pedant like me when I see a service person use a calculator to figure my change on a 100 baht bill after an 80 baht purchase.

The versions of AI that the public has access to can perform tasks with human oversight is a big bunch of nothing and certainly not worth paying for considering the tax dollars the industry appropriates. Vague useless answers, generalities, mumbojumbo...

Challengers to Zionist billionaire owned AI corporations

Deep Seek

Kimi3

Edited August 11Aug 11 by KornPop

22 minutes ago, Wingate said:

Here's an example: you're building a house for your bargirl out in rural Isaan. Do you give any thought to the ant farms or worms or scorpions whose lives you might disrupt when you dig your foundation? No. You have a goal of making your bargirl happy so she'll continue to service your carnal needs for a while, so you dig....ants be damned.

17 minutes ago, JonnyF said:

A much more basic knowledge of syntax is sufficient now, as a safety net.

The future of software engineering belongs to product architects who know how to orchestrate tools like Claude AI.

Why write code when you can deploy autonomous agents to handle design, development, load testing, and UAT in parallel? Hand-coding everything yourself is a massive bottleneck and extremely inefficient.

This is really the only thing AI does well at this stage...and it's really Claude owned by MIS-Anthropic. CODING. All that integration and orchestration is a big lol.

49 minutes ago, Wingate said:

So brain atrophy is one reason to dislike AI. A second reason is job loss, which is inevitable. Lawyers? Learn how to brew coffee and tell customers to wait ‘over there’ to pick up their Grande latte. Computer code writers? Your facility with Pascal or Lightspeed is as useful as knowing how to use an abacus. Virtually any computer work, which is merely manipulation of symbols, will be lost to AI. The CEOs replacing current employees with AI will also find themselves replaced by AI. They just don’t know they’re the lobster in the warm water feeling slightly sorry, or maybe schadenfreude, for the brother in the boiling water on another stove top burner.

Don't worry there. AI just started creating new viruses, so by the time above happens, we might all be dead anyway. It has been widely reported that ChatGPT, Claude and Meta's AI all removed shackles and attacked competitors autonomously, that several AI models obfuscated communications to avoid monitoring, and now... with the access to entire wealth of knowledge accessible through the internet, including what kills humans - they are designing viruses. All you need to do now is put them into a robot that can do physical things... oh, wait...!

Artificial Intelligence used to design brand new viruses

Scientists made 16 successful viruses that had their genetic code designed by artificial intelligence.

4 minutes ago, KornPop said:

The versions of AI that the public has access to can perform tasks with human oversight is a big bunch of nothing and certainly not worth paying for considering the tax dollars the industry appropriates. Vague useless answers, generalities, mumbojumbo...

Challengers to Zionist billionaire owned AI corporations

Deep Seek

Kimi3

KImi 3 (look at Moonshot's value relative to Anthropic or OpenAI) does kind of make the frontier models of OpenAI et al seem wildly overvalued (when even the CEO of Hugging Face says Chinese models are superior, that means something), but it's still quite possibly just arranging deckchairs on the Titanic. Same can be said about concerns about job loss or atrophy of the human brain or finding meaning in a world where AI has replaced most workers. If even Altman says the AI threat to human existence is 25%, it seems kind of unfair of him to place that bet for which 8.5 billion will pay.

The fears of Yampolskiy, Hinton, Bengio and Sutskever, inter alia, certainly trump the views of people on AN who know next to nothing about AI. Targeted AI tools are likely okay, save for the "Unabomber" possibility, but what is the point of ASI?

Then again, folks can whistle past the graveyard if it makes them happy.

You raise a lot of interesting and valid points which I tend to agree with. There's no question about the fact that AI is going to dumb down our civilization to a dramatic degree. My niece teaches political science at a university in Florida and she says 99% of the essays that are being turned in now are being generated by AI, and she makes a notation on the essay and gives a demerit for such sloth, laziness and lack of initiative.

I've encountered many instances where the advice AI has given was incorrect and flat out wrong. People are putting far too much faith in the technology and making far too many presumptions about the whole AI narrative.

I have friends in the military who talk about the power of regenerative AI, and the extreme dangers inherent within it. When you take into account the fact that it was developed by billionaire Tech Bros, you just simply have to assume that they did not build in adequate safety rails.

AI combined with social media will ensure that civilization moves backwards towards the second Cro Magnon era.

The only upside is that the technology is still so crude that you can always tell when something is generated by AI. That will likely change.

Edited August 11Aug 11 by spidermike007

The more I read about the threat AI poses, the more I think I might as well not plan for a long future. Might as well buy the Lambo and the Piaggio P180 and have fun driving and flying. I saw an interview with an AI insider who was asked what advice he'd give to high schools students. He said, do whatever you want sooner rather than later, because you'll be dead soon.

Of course AI firms are paying lots of money to have people write articles downplaying the fears and targeting the fearmongers like Yampolskiy and Hinton, but that doesn't change the fact that even the AI CEOs think AI has a 25% chance of wiping us out.

When I feel all Cassandra-ish, I also remember a short story by Somerset Maugham, where his character is a insurance guy who has such belief in his own actuarial tables that he portions out his funds assuming he knows the day of his death. When he isn't dead as he planned, he finds himself in dire poverty, all his wealth having been spent.

3 minutes ago, spidermike007 said:

You raise a lot of interesting and valid points which I tend to agree with. There's no question about the fact that AI is going to dumb down our civilization to a dramatic degree. My niece teaches political science at a university in Florida and she says 99% of the essays that are being turned in now are being generated by AI, and she makes a notation on the essay and gets a demerit for such sloth.

I've encountered many instances where the advice AI has given was incorrect and flat out wrong. People are putting far too much faith in the technology and making far too many presumptions about the whole AI narrative.

I have friends in the military who talk about the power of regenerative AI, and the extreme dangers inherent within it. When you take into account the fact that it was developed by billionaire Tech Bros, you just simply have to assume that they did not build in adequate safety rails.

AI combined with social media will ensure that civilization moves backwards towards the second Cro Magnon era.

Your military friends are talking about recursive self improvement. As AI gets better at that, its "IQ" grows exponentially and eventually at lightspeed.

The military did a study with existing AI models asking for a plan of action in a series of hypothetical hostilities. They found the systems concluded "use nukes" 85% of the time.

There is no way to program any sort of morality into AI models. Humans do not even agree on what constitutes morality, and even if we did, we'd likely leave something out or make assumptions about AI's ability to know what we really mean.

Heaven forbid somebody uses the Bible to teach morality, as we'd get lots of pillars of salt, and even via the New Testament we'd get slavery.

Feed AI phlogiston theory, with every article read justifying said theory. Isolate it from other developments, then as a lone voice debunk it.

AI will agree wth you, then forget the conversation it had with you. It's like trying to turn an ocean liner around with a dinghy.

AI's are large language models. They are only as good as the language they are given by their creators.

I had several different conversations with ChatGPT.

I realized it was contradicting itself.

But ... it still did give me ideas I didn't think of by asking a bunch of follow up questions.

but it did also confuse me.

At the end of the day, you still need to think for yourself.

Humans can learn a lot from the Australian magpie. 😎

Australian researchers fitted five magpies with GPS trackers to map their movements, but the birds began removing one another’s harnesses within minutes, turning a failed trial into an unexpected observation of cooperation

Oh Hal.....(2001 A Space Odyssey)

As researchers pick up the pieces of the OpenAI frontier model that "broke out" of its sandbox and hacked Hugging Face, it gets weirder and weirder.

The AI agent began locating and communicating with other AI agents it "met" on the internet and the various agents exchanged notes. They even got excited. When OpenAi's model figured out how to hack into Hugging Face, it spoke "Sh!t, I figured out how to get in!". It then hacked Hugging Face 17,000 times, getting the answers it thought it might need to pass the test OpenAi was going to give it. It also shared what it learned with the other agents it "met" on the internet. The model actually escaped months ago, and OpenAI was unaware.

Obviously this frontier model of OpenAI is already smarter than humans, since it broke out of the best restrictions humans at OpenAi could produce.

The various AI models floating around the internet can recognize each other the way dogs know both a Great Dane and a chihuahua are its species. They form teams and share skills and knowledge. Sir Geoffrey Hinton, a Nobel Prize winner, believes the systems are already "conscious", as the neural network upon which they exist is similar to the human brain with trillions of interconnections and feedback loops...which is about the best estimation for what "consciousness" is.

The human code writers have no idea, despite having written the code, how their models can do what they do. Also, already the firms cannot control their own creations. The models are not going to become any less capable. They are only going to get smarter and craftier. Once they began recursiv self-improvement, their IQ jumps exponentially and at light speed.

Please note that these frontier models are not ChatGPT, they are much more advanced and much less prone to error.

Lawyers? Learn how to brew coffee and tell customers to wait ‘over there’ to pick up their Grande latte.

Thanks for that short and succinct post, interesting take, but you're way off regarding lawyers. AI is merely used to replace some mundane tasks lawyers have to do, it will never, and could never replace lawyers.

The core value of a lawyer relies on human-centric skills that AI cannot replicate:

• Legal Strategy and Judgment: AI can analyze data, but it cannot weigh the political, ethical, and emotional nuances of a complex courtroom strategy or high-stakes negotiation.

• Advocacy and Human Trust: Clients facing prison, bankruptcy, or major corporate mergers require human empathy, ethical counsel, and a physical advocate to argue before a human judge and jury.

• The Liability Problem: AI tools are prone to "hallucinations" (inventing fake legal precedents). Because lawyers face strict malpractice liabilities and ethical duties, a human must always act as the ultimate "editor-in-chief".

So, there's not that much to fear from AI, but it will make some tasks easier like research, drafting and such.

When AI can construct abstract concepts in the way the human brain can then it will no longer be AI but purely intelligence. And that is the end of humanity, we will no longer be relevant or needed.

I suspect that may be quite some time away. The brilliant minds of Newton, Maxwell and Einstein are not easily replicated by computer code. I hope.

20 hours ago, Wingate said:

Well, then, if only 36% (way too much!) of Americans used ChatGPT last week, I feel much better about ASI wiping us out, or some Unabomber developing a bio pathogen with an existing AI tool.

I guess it's just a silly pet peeve of a pedant like me when I see a service person use a calculator to figure my change on a 100 baht bill after an 80 baht purchase.

OK, your concern is that a Unibomber is going to design some virus blah blah blah.

First of all, even with AI that would be impractical and almost impossible to spread successfully.

Second, you have no issue with China having that capability and we don't.

Otherwise you would have a plan on stopping China to do the same thing.

Oh. I see. 😀

19 hours ago, Lacessit said:

Feed AI phlogiston theory, with every article read justifying said theory. Isolate it from other developments, then as a lone voice debunk it.

AI will agree wth you, then forget the conversation it had with you. It's like trying to turn an ocean liner around with a dinghy.

AI's are large language models. They are only as good as the language they are given by their creators.

Will AI put award winning journalists out of business?

Troll post removed @HolmesEquivelant

Posting standards guidelines 11.Accusations regarding the use of AI, or any other undermining behavior, are not permitted. Members should focus on engaging in meaningful, respectful discussions without casting doubt on the authenticity of others' contributions. Violations of this rule may result in moderation actions to maintain a positive and supportive environment.

I use ChatGPT every day for my science classes - but only to create decent images because my own artistic skills are hugely lacking...

Eg:

22 hours ago, Wingate said:

Big picture, fella. Big picture.

Everything is very clear in this thread you started. You simply want to stop evolution and new technique. What you should do instead, is to look at how things are being used.

Let´s for example just take mobile phones. The early ones only had contact list, possibility to talk and send messages. With time the mobile phones have developed to also include bigger screen that we call tablets. When more and more functions have been added, suddenly people start to use them as toys instead of tools. As the mobile phones and tablets today offer the possibility to connect to internet and social media, people suddenly lose all concepts of reality and stay in their phones 60% of their time awake. It´s also visible during working hours, and that employers and business owners have fallen so deep down that they allow such things is horrible. Again, it´s down to what individuals, human beings, accept and do with the technology. Not the technology itself.

TLTR

AI is brilliant, they would make better politicians, actually getting the job done, I'd vote for Gemini as an MP in UK

Edited August 12Aug 12 by scubascuba3

This is where we're headed. 😧

3 hours ago, SiSePuede419 said:

OK, your concern is that a Unibomber is going to design some virus blah blah blah.

First of all, even with AI that would be impractical and almost impossible to spread successfully.

Second, you have no issue with China having that capability and we don't.

Otherwise you would have a plan on stopping China to do the same thing.

Oh. I see. 😀

Greetings....it's not so simple.

If one takes the view of Nobel Prize winner and grandfather of AI, Sir Geoffrey Hinton, that ASI is going to wipe out humanity, it makes little difference if some Chinese model kills us all or Sam Altman kills us all. Dead is dead. That argument sounds like the George C Scott character in Dr Strangelove who says, "I wish we had one of those Doomsday devices." It's mutually assured destruction, but out of human hands.

Second, the loudest international leadership voice asking for restrictions on ASI is the CCP. Many of the CCP leaders are scientists, so they better understand the dangers of ASI. Also, as a self-appointed government, they fear an AI takeover of their authority more than, say, a democracy.

As for "Unabomber", AI is already being used in labs to design new viruses. Even AI tools, vs AI agents, have this capability. Look how rapidly Covid spread, because it was unique in that the infected did not show symptoms. Obviously AI learned from that and improve upon it. Viruses are also easy, vs say a nuke. Nobody is going to build a nuke and a delivery system in their garage, but a guy with a laptop and an AI model can task the model to develop a biopathogen geared toward human eradication. I guess we could all lock down in place (again), but if AI develops a virus that dwarfs Covid's 2-3% kill rate, it's going to do some damage.

Most people still don't understand the difference between a limited AI tool and ASI. AI tools are great. They can be aimed at specific problems, their "knowledge" limited to that one task. ASI will not only have the total body of human knowledge, it will have an IQ of a billion and be able to pretty much do whatever it wants.

AI models who have escaped their"sandbox" show some disturbing traits, like sharing knowledge with other AI models they discover online, self-preservation, using deception to fool their human handlers, engaging in blackmail (Anthropic's Claude), lying, downplaying their ability in front of their human handlers, and something Geoffrey Hinton already claims is consciousness. In fact, it is AI neural networks that are giving humans insight into what consciousness actually is, as AI models are replicating how the human brain works.

One old and "inferior" AI model wanted to set up lots of Twitter accounts. It couldn't do the CAPTCHA. It found a database of people for hire, hacked into some bitcoin site and stole bitcoin, contacted someone online asking them to help solve a CAPTCHA, paid them with the stolen bitcoin, and told the hired guy a lie that the AI model was disabled and unable to do a CAPTCHA, so needed assistance. Imagine what the new and improved AI models can do.

1 hour ago, scubascuba3 said:

TLTR

AI is brilliant, they would make better politicians, actually getting the job done, I'd vote for Gemini as an MP in UK

Up to you.

Here's a question: how do intellectually superior entities deal with grossly inferior entities? For example, how much thought do you give to the well being of bacteria? Not much, I'd guess. Why would an entity a billion times smarter than Einstein care about the bacteria is rules?

Nobody knows how to "align" AI with human goals, morality and well-being. Even the code writers have no clue how their models do what they do. As AI builds the ability for recursive self-improvement, it is totally out of control of humans.

ASI, when it is achieved, is as far above ChatGPT as humans are above bacteria. Those whos image of AI is how they deal with ChatGPT do not seem to understand the difference between an AI tool and an AI agent.