The love of Elvis is the root of all money

 

here's a decent video on topic of "Will digital intelligence replace biological intelligence?", art is also mentioned there in Q&A part:

 

Yes, I think we all do atm.
But so far, the focus has been exclusively on pattern recognition, safety alignment and enlarging context window. These GPTs do not have any real logical capabilities, but derive reasoning from patterns. If you want to make a humanizing comparison for a better understanding, you could say: GPT-4, in its current state, is a lazy brain on autopilot. It answers without having thought about the question, based on learned patterns. Just as most people answer the question "How are you?" with "Fine, how are you?" without having thought about the question or answer. If you take this fact into account, it is very impressive actually, because not a single person I know whose brain is on auto-pilot is capable of responding in such a complex way as GPT-4. But that's all water under the bridge. Even Gemini 1.5 Pro, which is not even available to the public yet, is already old news. The latest studies are already much further along in terms of context window scalability, latency and (real!) active logic. The next few months will be extremely interesting, I can assure you.

Of course not. See answer above.
Just wait for it. The generation coming out this year will not achieve this either, but it will lay the foundations for it.

Companies have a problem filling IT positions. Have a look at the big tech companies for open positions ;)
People simply don't want a job with the most boring task imaginable when they've just graduated from an elite university and are $150K poorer. Why do you think there is so much focus on communication and coding in the development of AI?
Take an AI that is able to communicate with humans, to code, to grasp real, complex problem contexts in milliseconds and to solve them independently from human guidance by coupling this AI with an autonomous intelligent agent. This is not sci-fi, but existing state-of-the-art tech. And it solves one of the biggest current problems of big-tech companies (and of course the IT problems of smaller companies). The CEOs even say this publicly, "Save yourselves the trouble of learning to program, we don't need you."

As I said, the current public available big-tech models are long out of date. OpenAI's GPT-4 is actually from 2022. GPT-5 is just around the corner, as is Gemini 1.5 Pro and some new competitors. And these aren't what's new, it's what they've done in the last years. Transformers are no longer the big thing at all. It's already showing its limits. The focus of research and development has long been on other, much more powerful methods and technologies. And creative work is not and never has been the focus of development at all. Because creative work can be broken down to simple logic and conceptual understanding. What you are seeing in relation to semi-"artistic" image and video generation is just a side result of training AI a physical world concept before moving onto the next phase. These are just products that serve the sole purpose of generating financial income and collect user input data for the further development of the actual objective.

 

Last edited: Mar 7, 2024

trillion-bucks question nowadays is - for human-like intelligence, do you need more source data or more connection patterns?
because if a dog has twice as many brain synapses than a cat, and isn't twice as smart, or elephant having more complex brain network than human etc.. what kind of special abilities people think they have AI couldn't have...

 

I think you've taken my IT reference too literally but I get your point and actually I'm ready to swap an AI for an over stressed, very bored IT staff member. It's a mechanical repetitive job, that companies still struggle to run well.

I meant that so many companies are incapable of upgrading or even running efficient systems in IT, which is kids stuff compared to Projects I work on.

This myth CEO's are being sold that on large super complex projects they can just 'plug in AI and start removing people and costs' is a fantasy.

 

Could we have this statement in more direct details?

"new political" - maybe, dictature instead elections?
"social solutions" - social rating like in China?

 

What do you see as the ultimate use case for the main AI companies? I read different targets being speculated about, and I hear different paths, many extremely expensive. And a lot of paranoid speculation about secret agendas.

I feel with AI there's a lot of theoretical potential that sounds astounding/crazy, and then someone else pops up and says 'yeah it'll cost 30 trillion to develop'.

So many conversations ignore real world use and execution and just focus on outlandish endgames that needing outlandish development scenarios. I think this cognitive dissonance is why some call this another tech bubble, and i kind of agree after working on complex digital projects for 30+ years.

 

Last edited by a moderator: Mar 7, 2024

Just a good example.

Sure, that's not what I meant. Of course, it won't replace the superstars of the respective disciplines and their assistants who work on the innovative and exciting large-scale projects, but the thousands and thousands of jobs that take care of regular, everyday IT work.

The implicit problem with this is that not everyone can become a superstar in their discipline. Not everyone has innovative ideas. Not everyone has a high level of education and skills or can even achieve them.

 

They are not very radical or new are they?

Personally we need to move past party politics, break up corporations, stop relying on corrupt human power hierarchies and start utilising people who can actually solve social problems and serve the peoples needs, not their own. Could be AI systems running this, I have no idea. I haven't seen a serious alternative yet but the outdated 'isms all need to go, we need new ideas and models.

 

Last edited by a moderator: Mar 7, 2024

Yeah, I'd say we are all in agreement, this is the fundamental issue that's needs solving.

Honestly, i would like us to solve it. If they can actually integrate the tech, 90% percent of the jobs that will be 'optimised' are soul sucking corporate jobs only benefiting the rich. Even mine is and i might survive. I'd leave tomorrow if our bills were covered.

Maybe its time we roll the dice because this world is already horrifically unfair and broken.

(oh and i left school and started working at 15, i'm basically an idiot peasant who got lucky)

 

Last edited by a moderator: Mar 7, 2024

a thing that has no purpose, soon fails or disappears. Humans are heirarchial, we exist within a pecking order, its the way we are wired. The systems of distribution (economics) exist because they serve this understanding.
should this understanding change, these systems must change or fail.

they become a thing with no purpose.

What then replaces them, has always been the difficult question.

 

maximizing profit, no matter what exactly it means, not even sure if money is the only profit item to consider,
right now it's quite remarkable only 7 (tech) companies basically dominate U.S. economy growth, for ex. solely nVidia company being equal to entire Canada's GDP

 

"Personally we need to move past party politics"
we already passed that in a medieval age...are you a fan of P. Kropotkin anarchy?

"break up corporations"
they have money and some even have private armies

"Maybe AI can uncover these new models?"
exactly the corporations make investments into AI...

 

I don't know if there are "secret agendas". If there are, they're at least secret enough that I don't know about them. I mean, seriously: There are enough openly communicated agendas by various influential people that are already dystopian enough for my taste, or have so many negative utopian elements, that it's actually unnecessary to invent conspiracy theories. So it's not that there's a lack of new societal concepts or ideas for the real world application of general artificial intelligence. I think it's more that most people imagine utopias as something other than the Huxley-esque visions of the future currently on offer by some billionaires. We need more utopias!

The ultimate use case... ?
Man, I don't know if an "ultimate use case" even exists as a realistically achievable goal, and I think it heavily depends on whose perspective you're using.
For AI company investors, CEOs, etc, it's probably about dividends and market dominance... Getting a foot in everyone's door, creating dependencies, dominating. From the small private end user to states, military, etc. I don't know if it's going to be as toxic as Broadcom, for example, but who knows.... Economics as usual. Should the ultimate goal of creating an ASI ever be achieved, imagine the incredible power that whoever owns it would have in their hands.
Anyway...
My guess is that the people who really have both hands in the AI development have exactly the goal that is openly communicated: The creation of an ASI for the solution of mankind's unsolved problems. Simply because they want to know if they can do it. Leaving aside questions of feasibility, this is - IMO - a noble and well-intentioned goal. But as a wise man once said, whose name I can't remember:

"The road to hell is paved with good intentions"

Under my tech nerd hat, I watch these very exciting times with a permanent grin, happily anticipating the bright future, but when I take off my hat, the worry lines that dig deep into my forehead become visible.

Yes, and it was not even ten years ago that experts were predicting that our generation would not live to see the advent of text-to-video. Apparently these experts were wrong.

Things evolve. The brightest minds on our planet are working to solve these very problems. Increasing scalability, increasing efficiency, reducing energy requirements and costs. What costs 30 trillion today may cost "only" 30 billion in two years. Or less, who knows? Why? Because the approach might be completely different. It's not the first time that great things have been developed that seemed impossible. It is also normal for crazy amounts of money to be quoted as a prerequisite. If you set your sights high, you get more. An old story.

 

Last edited: Mar 8, 2024

In a historical context, it is now very difficult to have a hot war where people engage in combat, because all the main actors have nuclear overkill abilities and if it gets started it will get out of hand fast.

that is to give context to what is currently occurring in the world.
Asymetrical warfare is the rage, using the internet and media to influence the masses to sympathy or outrage by equal measure. - rather than 2 armies on a battlefield.

AI is great tool for information warfare, it can manipulate people without them realizing it because they will think they are receiving information from other people, not a computer. This is the sort of thing governments love, its a candy store they can't resist.

Ai is also low budget(by military standards) and scalable, so all the big 5 can participate to whatever extent they wish. - really any country that can mount the necessary hardware and steal the talent can join in. much easier than figuring out how to make plutonium.

A long time ago in a land that no longer exists, many nations, but primarily two, participated in a thing called an Arms Race, each trying to outdo the other with size of their bombs, as a young lad I would get the afternoon newspaper from the mailbox, and read the headlines as i carried it back to the house, and it was full of megatons, and hydrogen bombs, and ICBMs and mr. kruschev. Eventually they achieved mutual assured destruction, which is referred to as MAD, because this acronym truly fits the mindset.
and then they all sort of gathered around like Sumo wrestlers with hands on each others shoulders and danced around for a half a century or so no one able to use this thing they had struggled mightily to create.

This, the AI hullabaloo, is as much an arms race as it is about "commerce" or "surveillance by assistance", or "software upgrades" to make "Software more user friendly".

Every great project has an outwardly facing reason, and inwardly facing reason. aka a cover story. Bluntly this is an arms race at its core, being dressed up as whatever fantasy the public will beleive. AI dream girl, software automation, a 2 day work week, whatever sells the public.

AI is just the latest arms race, and quite possibly the last race for the human race.

 

@Sinus Well ..


"One' man's dystopia's the other one's utopia."

Lois (2024)

 

@Lois Lane This is profound.
I'm naive. Why can'twe just be totgether

 

Imagine ich und du, Ich tue
Ich denke Tag und Nacht an dich, it's only right
An das Mädchen denken, das du liebst, and hold her tight
So happy totgether…

 

Okay, let's assume that things turn out as predicted and a large proportion of work is performed by autonomous AIs. Then a new value must be created to justify payment, because currently the monetary value of a person is measured by their work performance. There are already various concepts for a basic income - either unconditional or conditional. Well, I would say that a basic income is not a bad idea in itself and the right path is probably in the middle.
Presumably this concept already exists, and some clever mind has thought it through much further than I even could... however....

Here's my take on it:

• Link a basic income to the UN Charter of Human Rights and the constitutions of nations, provided they are in full compliance with human rights. Full compliance with universal human rights is the condition for receiving a basic income.
• Create an international trust fund into which states that have signed the UN Human Rights Charter and whose constitutions are in full compliance with it, and corporations, pay a sum X based on their real economic performance.
• Pay every citizen whose country pays into the trust fund a sum X as a basic income. This sum should be the same for everyone, regardless of the country's economic performance. The sum has to be staggered and must be very well thought out in order to avoid over- or underpopulation.. A possible scale, for example: $2000 for an adult, $1000 for the first child, $500 for the second child, $250 for the third child, etc. When the first child reaches adulthood and receives $2000, the second child becomes the first and receives $1000.
• Additional income can be earned through innovation, self-employment and social commitment. Additional earnings are taxed and benefit the fund.
• The fund must hold a certain reserve, based on the number of recipients, for unforeseen events; every additional dollar that exceeds the reserve is distributed among the recipients of the basic income.

I'm sure the concept has many flaws and pitfalls. But this is the best I can do with a few minutes of thought.

So true!

 

I think world has to rethink meaning of money and its impact on society,
nowadays in economical way, people are supposed to get most money from their work time,
and social status is considered by wealth (having car, house etc...),

but if all that isn't relevant anymore, what other values will define our society?
if universal income results in people not being obsessed about having jobs, maybe they'll be able to devote time to something more meaningful, like family, relationships, personal growth (but not driven by monetary gains?)

it's a bit strange AI companies are desperately pushing into creative jobs as if they were trying to destroy art values entirely,
while on the other hand AI companies pretend all other work areas and jobs will benefit from added "cooperation" of AI and humans?