Page 4 of 24

Re: AI danger

Posted: Sat Jun 14, 2025 11:51 pm
by iambiguous
As with so many things that can impact the thoughts we think, the emotions we feel and the behaviors we choose, there's the part where we grapple with AI philosophically and the part where it either benefits or is a detriment to us for all practical purposes existentially.

Then the part where people ask AI questions in regard to things it can't possibly have any real knowledge of. Like, say, sex?

"Yo, AI, is going down on someone sexually a reasonable behavior?"

In other words, we need AI to reach the point where it becomes as sophisticated as Gigolo Joe in A.I.

Joe not only has sex, he is considered to be particularly good at it. Also, he is aware of himself as a Mecha.

Then the part where A.I. grapples with the morality of human sexual acts...?

Re: AI danger

Posted: Sun Jun 15, 2025 12:01 am
by accelafine
Of course it depends on what is being defined as 'intelligence'. Most humans can't answer the question 'why is the sky blue?' Ask AI anything and it will generally give you a pretty accurate answer. By that definition it's 'more intelligent' than the average human now :lol:
Personally I think it will simply continue to be very useful in many fields. I think that so-called 'superintelligence' is a fantasy concept with no real meaning or possibility of definition.

Re: AI danger

Posted: Sun Jun 15, 2025 12:06 am
by accelafine
So you can stop fretting Gary :D

Re: AI danger

Posted: Sun Jun 15, 2025 7:10 am
by Fairy
JamesSisyphus wrote: Sat Jun 14, 2025 9:46 pm
Fairy wrote: Sat Jun 14, 2025 10:04 am Humans are not intelligent.
This is true.

Present AI is, by definition, not intelligent. It is reliant upon statistical modelling, without inherent understanding.

People assume that “it’s just a question of computational power”, but this is wrong. That computational power has to be correctly directed.

Humans somehow balance models with actual understanding. We have no idea how to implement understanding, or to recreate the balancing act artificially.
Maybe there are many possibilities in how AI plays out, many scripts or timelines already available, and the one that plays out depends on you, since it’s your dream.

The way it plays out will depend on the thought system you foster. If your internal world is murderous then AI will have a destructive intent on the world, but if your internal world is kind and loving, then that’s how AI will appear in your timeline.

That’s how quantum mechanics work don’t they? All possibilities are there and you will experience the one you align with.

Re: AI danger

Posted: Sun Jun 15, 2025 11:43 am
by JamesSisyphus
accelafine wrote: Sun Jun 15, 2025 12:01 am Of course it depends on what is being defined as 'intelligence'. Most humans can't answer the question 'why is the sky blue?' Ask AI anything and it will generally give you a pretty accurate answer. By that definition it's 'more intelligent' than the average human now :lol:
Personally I think it will simply continue to be very useful in many fields. I think that so-called 'superintelligence' is a fantasy concept with no real meaning or possibility of definition.
It could only answer the question because it has already been answered by humans. If you ask it a question that has not been answered by humans (and encapsulated within the database), the LLM will hallucinate an answer.


Quantum mechanics is irrelevant.

Re: AI danger

Posted: Sun Jun 15, 2025 12:38 pm
by accelafine
JamesSisyphus wrote: Sun Jun 15, 2025 11:43 am
accelafine wrote: Sun Jun 15, 2025 12:01 am Of course it depends on what is being defined as 'intelligence'. Most humans can't answer the question 'why is the sky blue?' Ask AI anything and it will generally give you a pretty accurate answer. By that definition it's 'more intelligent' than the average human now :lol:
Personally I think it will simply continue to be very useful in many fields. I think that so-called 'superintelligence' is a fantasy concept with no real meaning or possibility of definition.
It could only answer the question because it has already been answered by humans. If you ask it a question that has not been answered by humans (and encapsulated within the database), the LLM will hallucinate an answer.


Quantum mechanics is irrelevant.
I didn't mention quantum mechanics. But yes, I'm sceptical that it will ever be able to answer a question that has never been answered by any human and which humans are incapable of answering.

Here's someone with a different approach from the 'we're all doomed' ones.

https://www.youtube.com/watch?v=Y1QOf6HEbHQ&t=1680s

Re: AI danger

Posted: Sun Jun 15, 2025 7:25 pm
by Gary Childress
accelafine wrote: Sun Jun 15, 2025 12:38 pm
JamesSisyphus wrote: Sun Jun 15, 2025 11:43 am
accelafine wrote: Sun Jun 15, 2025 12:01 am Of course it depends on what is being defined as 'intelligence'. Most humans can't answer the question 'why is the sky blue?' Ask AI anything and it will generally give you a pretty accurate answer. By that definition it's 'more intelligent' than the average human now :lol:
Personally I think it will simply continue to be very useful in many fields. I think that so-called 'superintelligence' is a fantasy concept with no real meaning or possibility of definition.
It could only answer the question because it has already been answered by humans. If you ask it a question that has not been answered by humans (and encapsulated within the database), the LLM will hallucinate an answer.


Quantum mechanics is irrelevant.
I didn't mention quantum mechanics. But yes, I'm sceptical that it will ever be able to answer a question that has never been answered by any human and which humans are incapable of answering.

Here's someone with a different approach from the 'we're all doomed' ones.

https://www.youtube.com/watch?v=Y1QOf6HEbHQ&t=1680s
I wonder if he works for the private sector. People in the private sector will tell you what they're selling is great stuff, no matter what. I listened to NPR a while back regarding a conference to try to handle plastic waste through recycling to reduce production of it, and the industry guy was harping about all the "great opportunities" the conference offered but after some questioning, it turned out that he had nothing to offer other than increasing the production of plastic with the assumption that recycling would handle it eventually, even though much recyclable plastic now ends up in waste dumps regardless.

Re: AI danger

Posted: Sun Jun 15, 2025 7:49 pm
by accelafine
Gary Childress wrote: Sun Jun 15, 2025 7:25 pm
accelafine wrote: Sun Jun 15, 2025 12:38 pm
JamesSisyphus wrote: Sun Jun 15, 2025 11:43 am

It could only answer the question because it has already been answered by humans. If you ask it a question that has not been answered by humans (and encapsulated within the database), the LLM will hallucinate an answer.


Quantum mechanics is irrelevant.
I didn't mention quantum mechanics. But yes, I'm sceptical that it will ever be able to answer a question that has never been answered by any human and which humans are incapable of answering.

Here's someone with a different approach from the 'we're all doomed' ones.

https://www.youtube.com/watch?v=Y1QOf6HEbHQ&t=1680s
I wonder if he works for the private sector. People in the private sector will tell you what they're selling is great stuff, no matter what. I listened to NPR a while back regarding a conference to try to handle plastic waste through recycling to reduce production of it, and the industry guy was harping about all the "great opportunities" the conference offered but after some questioning, it turned out that he had nothing to offer other than increasing the production of plastic with the assumption that recycling would handle it eventually, even though much recyclable plastic now ends up in waste dumps regardless.
I didn't get that from him. I found him quite balanced and objective. There's no need to 'promote' it anyway. It's here to stay and we just have to sit back and wait for humans to do the wrong thing as usual.

Re: AI danger

Posted: Sun Jun 15, 2025 8:15 pm
by Gary Childress
accelafine wrote: Sun Jun 15, 2025 7:49 pm There's no need to 'promote' it anyway.
If you work for a private company, the CEO wants passionate people on his team. When he meets with investors he wants to be able to say that investment shares will skyrocket from all the hard work and effort they put into their product, be it plastics or AI. I've seen corporations at work on many levels. If there's profit in it, and it isn't outright illegal, most CEOs will find someone passionate about the product. Tobacco companies hire people who believe in "freedom of choice" and whatever else smokers may need to think to consume their product, for example. Money seems to do a lot of corrupting. If some of the greatest AI people out there are concerned, then it sounds like good reason to me to take heed.

Re: AI danger

Posted: Sun Jun 15, 2025 9:51 pm
by accelafine
Gary Childress wrote: Sun Jun 15, 2025 8:15 pm
accelafine wrote: Sun Jun 15, 2025 7:49 pm There's no need to 'promote' it anyway.
If you work for a private company, the CEO wants passionate people on his team. When he meets with investors he wants to be able to say that investment shares will skyrocket from all the hard work and effort they put into their product, be it plastics or AI. I've seen corporations at work on many levels. If there's profit in it, and it isn't outright illegal, most CEOs will find someone passionate about the product. Tobacco companies hire people who believe in "freedom of choice" and whatever else smokers may need to think to consume their product, for example. Money seems to do a lot of corrupting. If some of the greatest AI people out there are concerned, then it sounds like good reason to me to take heed.
Don't worry. I know all about phony American bullshit. It's why we are in the mess we're in. 'Passionate' about hemorrhoid cream and nose clippers? Sure :roll:

I once applied for a job packing shelves in a supermarket for extra money. When asked at the interview, 'What is your ideal job?' I answered, 'theoretical physicist'. Needless to say I didn't get the job. How was I supposed to know I was meant to lie and say, 'I'm very passionate about baked beans and other canned products. I relish the opportunity to learn how shelf position is scientifically designed to maximise profit. My eye for detail will ensure that no baked bean can goes unnoticed and unsold. My passion for your company and desire to help it continue to grow is second only to my passion for following the share prices of supermarket chains on a daily basis. This is the job I have always dreamed of having and if you hire me I will give 110 percent!'

I realised that my acting skills and lack of self respect were never going to be at the level required by the cancer of American bullshit so I decided to work for myself.

:wink:

Re: AI danger

Posted: Mon Jun 16, 2025 6:18 am
by Walker
Insurance Commercials have changed in the USA.

https://www.youtube.com/watch?v=tKdULhHpA1Q

Re: AI danger

Posted: Mon Jun 16, 2025 3:14 pm
by Walker
Emu had a dream
https://www.youtube.com/watch?v=UTJ0zh7LuHM

Satirizing Horatio Alger sells security and eases the daily grind of the grey, bureaucratic cubicles where you better signal the proper politics, Buster. Gently mocking those with dreams beyond the status quo via humour, even ridiculous absurdity, is next level propaganda if one is to get serious and philosophical about what persuades these days.

Re: AI danger

Posted: Mon Jun 16, 2025 3:28 pm
by Walker
One of the clan is a very old man who started out stocking shelves. He would chat up the salesmen who had that store as a client. He ended up getting a sales job through one of those contacts, got a job representing a nationwide chain, rose through the ranks to executive, then landed the biggest account in the history of the big company. He retired on easy street. This is why America is the promised land, but it’s still a predatory universe with adaptability at the top of the food chain.

Doing that when AI is in charge will add another dimension to the maze, especially if AI is guided by equity for all and requires that big house to be split up into apartments.

Re: AI danger

Posted: Mon Jun 16, 2025 3:57 pm
by accelafine
'The promised land' :lol: This is why you are a so disliked. Empty, arrogant blowhards.

Re: AI danger

Posted: Mon Jun 16, 2025 4:09 pm
by Walker
accelafine wrote: Mon Jun 16, 2025 3:57 pm 'The promised land' :lol: This is why you are a so disliked. Empty, arrogant blowhards.
It's a form of punking, e.g., a military parade to celebrate the 250th anniversary of the military in USA, USA, the land of the free and the home of the brave ... and Trump's birthday.

Drives America haters nuts.

This is why "deplorable," was so embraced. Verbal judo.