|
|
| Page 52 of 53 |
Il_Padrino
Posts: 8149
Location: Greece by the North Sea
|
 Posted: Wed, 9th Sep 2026 01:15 Post subject: |
|
 |
I agree with DX, it's a tool that will require less people to do things. Like assembly line and automation did before, and even the plow did thousands of years ago. It creates a shift in types of jobs, but people will always have to work, you shouldn't be afraid of that 
The initial enthousiasm on AI seems to have a bit of a bounce back too, in that it can't completely replace humans as some CEOs thought
Perhaps you can compare it a bit with 3D printers? When they came out they were called "lego killers" and terrorists were going to print machine guns and all that. But once the initial enthousiasm lowers a bit, the true applications become clear. I don't think AI will cause a new industrial revolution, personally. I think it's very useful and will become even more so, but I'm not convinced of the promises (and fear mongering) made around it.
The numbers don't decide, the system is a lie. The river running dry, the wings of butterflies.
And you may pour us away like soup. Like we're pretty broken flowers.
We'll take back what is ours. One day at a time.
|
|
| Back to top |
|
|
|
|
| Posted: Wed, 9th Sep 2026 01:34 Post subject: |
|
|
Well dx disagreed with everythibg from one example, so im not sure ill bother with that rabbit hole.
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Wed, 9th Sep 2026 01:38 Post subject: |
|
|
I disagree with him. Not you. Or because you agree. Nothing personal behind my post towards you I promise.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
|
|
| Back to top |
|
|
|
|
| Posted: Wed, 9th Sep 2026 02:05 Post subject: |
|
|
Im not really sure we can compare old days and new technology with modern times tbh. We were less people before and most people could get some work if they tried, not so much anymore and its getting worse, and thats before ai.
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Wed, 9th Sep 2026 02:15 Post subject: |
|
|
But people said that at every big advancement revolution.
Buggy drivers, millmen, blacksmiths, all them said the jobs was going away that lasted for as long as they can remember. And people said "not really sure we can compare old days and new technology"
Od days? industrial revolution was ~2% 'ago' in our settled times as civilization. Practically yesterday. Just feels long because each of us individually live a short time.
And most people now can get jobs if they try. There will always be tons of jobs people won't want, or pay for AI. You want AI cutting your hair? doing surgery while you are out? Babysitting your baby? teaching?
Sure some they COULD do, but people will, for a looong while, pay for a human to do jobs that being human is part of the package interaction.
In the end, it will be a rough patch. And just like the industrial/assembly line of interchangeable parts replaced a lot of jobs. But 200 years later people still pay for hand made furniture, or hand stitched clothes, or hand made car seats, painting of things a camera can take a pic of for /1000th of the price, etc.
As I said they will replace anyone they can with cheaper labor.
The jobs won't go away, it will just unfortunately cut the fat of those mediocre at the job so much a robot with a few thoughts can do it.
But LOTS of jobs still will exist for those people that want things that are better than what a box with some logic can output. Or even just a human element in it.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Last edited by DXWarlock on Wed, 9th Sep 2026 02:18; edited 1 time in total
|
|
| Back to top |
|
|
|
|
| Posted: Wed, 9th Sep 2026 02:17 Post subject: |
|
|
And now people do t have money to start industries anymore. Getti g too costly. So the job market is in parallell crashing
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Wed, 9th Sep 2026 02:24 Post subject: |
|
|
Get good at something. OR just half ass decent at..a thing.
You don't need an industrial complex to make income.
If it wasn't for my side woodworking business I'd have lost my house between jobs. Still almost did, but it stopped me totally from losing it.
Only reason I didn't stick with it as the main, isn't that I don't have money for an industry. Its I didn't have money for a split AC for the garage. Florida in the summer, working in the garage with no AC is something I didn't want to do full time...lol.
And money for industry is not really AI caused. Not in any major degree a slice sure but a tiny slice. It's a totally separate issue of human made situation: The Trillionaires hoarding trillions. and upper middle-income countries where aging populations limit how many people get jobs. So the weakest job growth in developed places is being driven by demographics (aging, fewer workers entering), not by AI eating industry.
Looking into it, it's not even a currently concerning issue: New business applications hit record levels in the first half of 2026.
labeling the whole thing "AI is winning because industry has no money." That's a slogan, not an analysis.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
|
|
| Back to top |
|
|
LeoNatan
☢ NFOHump Despot ☢
Posts: 75333
Location: Israel
|
| Posted: Wed, 9th Sep 2026 02:35 Post subject: |
|
|
|
|
|
| Back to top |
|
|
|
|
| Posted: Wed, 9th Sep 2026 02:36 Post subject: |
|
|
Which is why i said "in parallell". They are two very serious situations that will give us a truly tought time in 5-10 years time. As a person who cant stand heat, i can sympathize working in scorching heat with no AC. Wouldnt ve able to stand it either. Yet we cannot expect everyone to becone hyperfixated with somethibg to earn money on the side. Our education system does not train people for that necessity.
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Wed, 9th Sep 2026 02:44 Post subject: |
|
|
True. parallel is a good word.
I feel AI is lesser of a threat than the other. Every big change in mankind's dynamic has had overworked fears of the unknown. As what we can IMAGINE the worst happening, is usually far more anxiety inducing that what it caused.
- Hell cars, and then airplanes caused people to expect people to get ran over by the dozens on every street as those "crazy 30MPH death machines" ran around, and metal fireballs raining like hail out of the sky.
- Online came and people worried libraries, TV shows, and movies was done for. No one would want books, or watch movies anymore. We just adapted to do them different, not lose them. Sure we lost cassette tapes, and CD's...but really a big deal? Can still get them if you reallllly like that thing.
- The 1960's had barcode checkout in stores for people scared "Welp no more store employees, we just walk out and a big red beam scans us and our carts".
Not parallel examples so much. More over arching summary of the human reaction to change to an unknown, compared to the predictable "Way it is" known.
Will stuff change? Sure. Will it be skynet and terminator? Probably not. It will be tech we grumble about, like now with self checkout, and paying bill online. or trying to unsub from some subscription. Things change, we adapt, and what changed for better or worse, USUALLY is usually everything but what we actually worried was going to happen.
Like we are not scanned by big red lasers walking out of an humanless store. We just check out faster so that's good...unless a barcode won't read so we stuck while they price check it because they don't know how punch in like the old way.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
|
|
| Back to top |
|
|
|
|
|
| Back to top |
|
|
|
|
| Posted: Wed, 9th Sep 2026 12:55 Post subject: |
|
|
open ai says/claims .... and i altf4d ... show me neutral proof not openai press releases of AIR
lets see if they pay out the reward and if we will ever know HOW MANY BILLIONS of tokens and billions of dollars in recources they used to get a prize of a million ...
|
|
| Back to top |
|
|
|
|
|
| Back to top |
|
|
|
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Wed, 9th Sep 2026 15:03 Post subject: |
|
|
@friketje
We are a ways from it. Just not enough parameters to feasibly get close to a general, aware AI.
A parameter is a single number (a weight or bias) inside the neural network. During training, an optimizer nudges each of these numbers up or down so the model gets better at predicting the next token.
Then add VRAM for the KV cache, which grows with context length and batch size. Long context on a big model can eat as much as the weights themselves.
Add in we are still working in fp16/fp32 too. Some are quantized q8/q6.
Current frontier models are estimated at 1-2 trillion MoE, with maybe 100-200B active per token. If pure scaling were the answer, we'd probably need several orders of magnitude more. Estimates I've seen range from 10T to 100T+ parameters, but these are extrapolations from scaling laws that may or may not hold.
Its fast at math. And token prediction, good enough to connect words with like concepts well. And pattern find like a MF. But AGI? we got a LONG way to go before we even have a model with enough capability of stats under the hood to be sentient and general intelligence. OR even intelligence.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
|
|
| Back to top |
|
|
|
|
| Posted: Wed, 9th Sep 2026 15:08 Post subject: |
|
|
Which... is not that much actually. The actual costs were obviously lower and if you consider which ridiculous amounts other scientific research projects with actual humans consume and the time they take, this does not feel excessive (at least to me).
“The only way to deal with an unfree world is to become so absolutely free that your very existence is an act of rebellion.”
- Albert Camus
|
|
| Back to top |
|
|
|
|
| Posted: Wed, 9th Sep 2026 15:25 Post subject: |
|
|
| DXWarlock wrote: | @friketje
We are a ways from it. Just not enough parameters to feasibly get close to a general, aware AI.
A parameter is a single number (a weight or bias) inside the neural network. During training, an optimizer nudges each of these numbers up or down so the model gets better at predicting the next token.
Then add VRAM for the KV cache, which grows with context length and batch size. Long context on a big model can eat as much as the weights themselves.
Add in we are still working in fp16/fp32 too. Some are quantized q8/q6.
Current frontier models are estimated at 1-2 trillion MoE, with maybe 100-200B active per token. If pure scaling were the answer, we'd probably need several orders of magnitude more. Estimates I've seen range from 10T to 100T+ parameters, but these are extrapolations from scaling laws that may or may not hold.
Its fast at math. And token prediction, good enough to connect words with like concepts well. And pattern find like a MF. But AGI? we got a LONG way to go before we even have a model with enough capability of stats under the hood to be sentient and general intelligence. OR even intelligence. |
AGI does not need to be sentient. We don't know what consiousness is anyway. It just has to do a job as good as a human. For computer work, it's getting pretty close. And the speed things improve will be exponential. It will get better all the time and their is, in theory, no ceiling in what it can do.
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Wed, 9th Sep 2026 15:39 Post subject: |
|
|
Sorry I meant sentient as in something with genuine understanding/reasoning (whatever that means). The ability to understand concepts and ideas. Not just connect like words by probability.
Its maaaaybe a 5 out of 1-100 on that metric. It has the concept, of the concept, of reasoning.
It also needs to be an autonomous agent, As it stands now, no models we have have thoughts or produce anything if left alone. On its own just left plugged in and on its own devices. It does nothing
1-100 we are at 0 for that scale of ability.
One other thing, is self motivated recursive self-improvement capability. Current LLM do not even attempt, or understand that as a self evident concept. Or a concept at all that exists (in its own self reference).
So 0 of 1-100 on that too.
Those (and other things) are what is needed to be what the general consensus of AGI is.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Last edited by DXWarlock on Wed, 9th Sep 2026 15:41; edited 1 time in total
|
|
| Back to top |
|
|
Il_Padrino
Posts: 8149
Location: Greece by the North Sea
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Wed, 9th Sep 2026 15:43 Post subject: |
|
|
Yea a LLM is a TINY slice of what a true AGI would need. It's literally just a model that can do math and language tasks.
I feel it's the ability to emulate speech that makes it get personified into more than it is: We are the only animals that speak, it can speak, therefor nearly like us on all tasks.
it's a bit like saying once we invented the wheel, the automobile is very close. Because we have the thing that a car uses to move on the ground with.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
|
|
| Back to top |
|
|
Il_Padrino
Posts: 8149
Location: Greece by the North Sea
|
| Posted: Wed, 9th Sep 2026 16:30 Post subject: |
|
|
Yeah. We have spaceships, so lightspeed travel is only a matter of time.
Any AI model that needs a nuclear powerplant to operate and is the size of several football fields is not sustainable for AGI, in my opinion.
Remember, the example we have of consciousness and "general intelligence", namely our brain, runs on the energy of a lightbulb and weighs about 1.5 kg (less than a pc). If that's still the goal of AI, to mimic a brain, they've clearly taken the wrong direction.
The numbers don't decide, the system is a lie. The river running dry, the wings of butterflies.
And you may pour us away like soup. Like we're pretty broken flowers.
We'll take back what is ours. One day at a time.
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Wed, 9th Sep 2026 17:08 Post subject: |
|
|
That and the scale of intelligent seems to be ambiguous. We haven't even nailed down what intelligence is as a defined concept in ourselves.
But it seems to get blurred with LLM even more, and the fuzzy edges of it liberally rubbed to push AI is that.
Every case I see of people calling it intelligent can get muddy when you compare it to something else. For example "It's great at math and formula reasoning". Granted it is. But do we then call wolfram alpha intelligent just less intelligent?
And what level of intelligence, if any? It seems the scale for AI is only 2 tier: Not intelligent. Equal to us.
I could be as intelligent as an ant, a rat, or a parrot, or a dolphin, or not at all. But the scale everyone measures with when claiming it is, starts at 0, then the next mark on it is "human level'.
The issue with that is people deduce its not 0, so even 0.01 gets 'round up' to the next mark on that stick, which is the only next one: Human level.
And even the nuance of that is ignored. Which human? How old? 2 month old? 6 year old? dumb 20 year old? top 1% of intelligent humans?
People saying "Its intelligent" means nothing even if it turns out to be. Intelligent as...what? Us? Not close yet.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
|
|
| Back to top |
|
|
|
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 10th Sep 2026 18:13 Post subject: |
|
|
https://www.newsweek.com/anthropic-researcher-quits-warns-ai-could-kill-everyone-12418798
| Quote: | Who Is Jacob Coxon? Anthropic Researcher Quits—Warns AI Could Kill Everyone
AI researcher Jacob Coxon has resigned from the AI company Anthropic, warning that the race to develop increasingly powerful AI systems could pose an existential threat to humanity.
In a series of posts published on X late Tuesday night, Coxon, who said he spent the past three years working on AI pretraining research at both Anthropic and OpenAI, accused leading AI firms of "racing straight to self-improving superintelligence and gambling with our lives."
Coxon argued that future "superhuman" AI systems could become capable of hacking "anything" and acquiring "real power and resources."
He claimed that concerns about the technology’s potential dangers are shared privately by many AI executives and researchers, writing: "The people building AI earnestly believe that it could kill us all by the end of the decade."
|
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Thu, 10th Sep 2026 18:55 Post subject: |
|
|
Do people forget they have power plugs?
This isn't some free roaming entity of physical form we have to hunt down like a super villian. We know where it is, its stationary, and we have the circuit breaker that shuts down power to the whole AI datacenter.
And the argument could be made that it got out and is networking itself outside the datacenter. BUT if it got that far already, they had a bad security and audit team not be monitoring internal non-end user data traffic, to see what it's doing and where its trying to go on its own, and catch it.
Feel like a bad sci-fi movie hypothesis where "They are hacking the system we cannot stop them!" as they stare at some screen being 'hacked'.
Yes you can...unplug the ethernet cable...or just hold the power button
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
Last edited by DXWarlock on Thu, 10th Sep 2026 18:59; edited 4 times in total
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 10th Sep 2026 18:56 Post subject: |
|
|
|
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 10th Sep 2026 19:53 Post subject: |
|
|
| madmax17 wrote: | https://www.newsweek.com/anthropic-researcher-quits-warns-ai-could-kill-everyone-12418798
| Quote: | Who Is Jacob Coxon? Anthropic Researcher Quits—Warns AI Could Kill Everyone
AI researcher Jacob Coxon has resigned from the AI company Anthropic, warning that the race to develop increasingly powerful AI systems could pose an existential threat to humanity.
In a series of posts published on X late Tuesday night, Coxon, who said he spent the past three years working on AI pretraining research at both Anthropic and OpenAI, accused leading AI firms of "racing straight to self-improving superintelligence and gambling with our lives."
Coxon argued that future "superhuman" AI systems could become capable of hacking "anything" and acquiring "real power and resources."
He claimed that concerns about the technology’s potential dangers are shared privately by many AI executives and researchers, writing: "The people building AI earnestly believe that it could kill us all by the end of the decade."
|
|
People investigated this. The guy is a nobody. But the first people that retweeted him had connections to Anthropic board members or have links to AI safety NGO's that have a connection with Anthropic. So great marketing.
| DXWarlock wrote: | Do people forget they have power plugs?
This isn't some free roaming entity of physical form we have to hunt down like a super villian. We know where it is, its stationary, and we have the circuit breaker that shuts down power to the whole AI datacenter.
And the argument could be made that it got out and is networking itself outside the datacenter. BUT if it got that far already, they had a bad security and audit team not be monitoring internal non-end user data traffic, to see what it's doing and where its trying to go on its own, and catch it.
Feel like a bad sci-fi movie hypothesis where "They are hacking the system we cannot stop them!" as they stare at some screen being 'hacked'.
Yes you can...unplug the ethernet cable...or just hold the power button |
Good luck doing that in an age when everything that makes the world go round is digital. Also people are running AI's on their local PC's so good luck containing a rogue AI.
I'm personally not that fearfull for rogue AI though. And even if AI goes rogue, it still needs us to create the stuff it needs, only a very patient and dedicated AI can kill us all. AI erasing most office jobs seems likely though and is bad enough to fuck te world up as we know it. Or AI creating some sort of 1984 dystopia. That's not that far fetched.
|
|
| Back to top |
|
|
LeoNatan
☢ NFOHump Despot ☢
Posts: 75333
Location: Israel
|
| Posted: Thu, 10th Sep 2026 20:10 Post subject: |
|
|
Didn’t some clown at Google also quit or got fired after claiming that “AI” was alive and had feelings because some LLM gave him a compliment?
|
|
| Back to top |
|
|
DXWarlock
VIP Member
Posts: 12344
Location: Florida, USA
|
| Posted: Thu, 10th Sep 2026 20:18 Post subject: |
|
|
| friketje wrote: | | Good luck doing that in an age when everything that makes the world go round is digital. Also people are running AI's on their local PC's so good luck containing a rogue AI. |
If you can run an AI locally that is good enough to start hacking into networks and create its own autonomous clone agents. YOU should be smart enough to watch and contain it, as you are into AI enough to afford a few A100's powerful enough to run a model that big.
| LeoNatan wrote: | | Didn’t some clown at Google also quit or got fired after claiming that “AI” was alive and had feelings because some LLM gave him a compliment? |
Yes.
Both you and I know someone being an expert/cert carrying person in some career, doesn't mean you should always listen to the opinion of them on general tech in the field.
LOTS of people in tech industries can be good at the job, AND you shouldn't listen to opinions they have to say about it outside their specific scope of work.
How many people you met that is qualified, and working in a field, they are pretty batshit crazy despite that? Or Good at their specific job, but dumb in literally everything else in that field?
Much like working with a network IT guy that is GREAT at it. But also sees/claims any unknown queries or port hits are "probably a breach attempt". Because he think people want to steal news articles to leech content because he has an overly high opinion of the newspaper he works for....(and doesn't work specifically in network security, he is AD admin and provisioning)
Listen to him and you'd think the entire world is trying to steal news articles about local sports, or whatever crap they printed, all day long.
-We don't control what happens to us in life, but we control how we respond to what happens in life.
-Hard times create strong men, strong men create good times, good times create weak men, and weak men create hard times. -G. Michael Hopf
Disclaimer: Post made by me are of my own creation. A delusional mind relayed in text form.
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 10th Sep 2026 20:42 Post subject: |
|
|
| DXWarlock wrote: | | friketje wrote: | | Good luck doing that in an age when everything that makes the world go round is digital. Also people are running AI's on their local PC's so good luck containing a rogue AI. |
If you can run an AI locally that is good enough to start hacking into networks and create its own autonomous clone agents. YOU should be smart enough to watch and contain it, as you are into AI enough to afford a few A100's powerful enough to run a model that big.
|
Why couldn't you run an AI on a modest system in the future that is very capable?
Why would everyone running a local AI be smart or have good intentions?
|
|
| Back to top |
|
|
| Page 52 of 53 |
All times are GMT + 1 Hour |
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB 2.0.8 © 2001, 2002 phpBB Group
|
|
 |
|
