AI - Cheating or just evolution?

[bxbrkrz]

[avasopht]

About the AGI hype: Truth is, we have no idea how close or far we are to AGI.

GPT and other LLMs are a lot like the Yoruba's Ífa, or China's I-Ching divination that could allow you to construct random sentences by effectively rolling dice. It's just a lot more sophisticated, but are not indications of sentience or "intelligence".

They could play a central or important role in an AGI, but it is only a small step towards it.

What's great about them, is that it's made the use of AI much more accessible to programmers and consumers. Anyone can use Bing (free) or another AI chat agent to gain obtain information or assist in an analysis.

---

When it comes to music, art and other creative fields, I don't like the idea of corporations replacing humans with AI, even if it's only a percentage of us. They will replace some of us, and already have to some degree.

Record labels will use AI to get around paying royalties by having AI create similar to a piece of music.

Maybe they'll use it to predict our response to trends and deviations to outsmart innovative artists.

It'll be hard to resist it, as others have suggested, if we cannot detect it.



It could also democratise music making by making it easier for bedroom producers to get the sound of live bands, orchestras, choirs and vocalists, so it's not all doom and gloom.



Only time will tell, but the likely outcomes are concerning.

You'll often hear that efficiencies and technological revolutions will just create new jobs and industries. But they actually haven't. Our productivity has increased at least a thousand-fold in the last 150 years, and in exchange, the cost of housing and living has not decreased one iota.


My area of work is building the pipelines that aggregate, manage and supply the data that would be used by analysts and AI. The things I've seen. Even before the hype of Chat GPT, we had a simple AI model replace expertise that previously required a human to have studied an obscure subject to the PhD level and become one of a handful of global experts over the course of at least 10 years of work experience. It was necessary because there is a growing shortage of the capability. But at the same time, it made that skill completely obsolete.

I'm only saying that to caution against underestimating their impact. Ignore the hype too, but don't ignore the potential impact.

[PhillipOrdonez]

lol that fallout vid is so

The guy can’t even fathom a hypothetical world where there’s “infinite abundance” and “intelligence explosion” without the shackles of neoliberalism and capitalism and “tHe mArkEt” ubi Of course the whole premise is totally wrong, agi is pure fantasy, thankfully

[PhillipOrdonez]

[avasopht]

The guy can’t even fathom a hypothetical world where there’s “infinite abundance” and “intelligence explosion” without the shackles of neoliberalism and capitalism and “tHe mArkEt”

We already have abundance. More food is thrown away than is needed to feed the hungry.

The problem is, as it has always been, the accumulation and concentration of power and control over abundant resources.

Of course, I wouldn't want to drag the discussion in another direction. But we became abundant with the industrial revolution. Everything we did to live can be done with a tiny fraction of human effort. That's why 80% of jobs are in the service industry (these are jobs that are largely unnecessary and only exist because of the things you mention ).

agi is pure fantasy, thankfully

Well, it all depends, ... what do you see as the big difference between current Artificial Intelligence methods and Artificial General Intelligence that makes AGI absolutely impossible now, and in the future?

Be mindful of the unconscious mind's tendency towards cognitive dissonance. AGI is scary. Defense mechanisms run a level or two deeper than our intellect.

---

Even if AGI were fantasy, the capabilities of AI today are already sufficiently disruptive. They're just a little overhyped. But nonetheless disruptive.

As I said, I saw it being used to replace a role that required more than a decade of experience by an expert (which was necessary because nobody studies that shit anymore, even if some of the richest companies would pay top dollar for it).

And more importantly, Artificial General Intelligence is not necessary for much of the impact expected (and feared) by AGI.




Generalized Intelligence is the idea that all of human intelligence can be explained by a single capability that generalizes to all problems. It's why the brain is believed to be able to adapt to physical damage by using other parts of the brain previously used for one function to fulfil the needs of another.




Artificial Generalized Intelligence is a theoretical inevitability (the only way it could not is if generalized intelligence is achieved in humans by magic that defies physics, maths and logic). The only thing that would prevent practical AGI is if it requires inordinate amounts of computational power.

There is nothing mystical about our thinking ability. It's just that it simulating biological processes involved in human general intelligence is prohibitively computationally expensive, and a coming up with a cleaner theoretical that is less computationally expensive will take considerable time.

[PhillipOrdonez]

The problem is, as it has always been, the accumulation and concentration of power and control over abundant resources.

yes. I am making fun of his inability to see beyond his nose, like most other people, when it comes to systems of governance. Mind you he's talking about INFINITE abundance. Currently there is abundance, it has been so for decades as you suggest, I remember the late futurist Jacques Fresco started talking about this long before the 80's

Well, it all depends, ... what do you see as the big difference between current Artificial Intelligence methods and Artificial General Intelligence that makes AGI absolutely impossible now, and in the future?

You've answered that question yourself here:

magic

Let me explain my thinking, which is simple, I don't have the big words or big knowledge in biology and physics and math and philosophy to put this into better terms: I think that being an organic being, made of smaller living things, may have a little something to do with it. you can call it magic, I don't know what to call it other than "being alive" :shrugs:

it requires inordinate amounts of computational power.

That too, thankfully. meaning it will be a waste of money (already is for most tech companies) and a waste of natural resources (already is, we need to tell everyone to stop using that crap)

But nonetheless disruptive.

No doubt, but it does not mean that it will stick and be more than a passing fad and become profitable. The generative aspect of machine learning is increasingly seen negatively by most regular people.

I am not saying that because it is useless crap and a fad, it is harmless. Quite the contrary I am advocating for the resistance to it and I think it starts by disspelling the doom and gloom that ASSUMES all they are saying about it is a fact that will happen without uncertainty. That leads to inaction and despair. I want us to fight against it instead.

[PhillipOrdonez]

Sam Altman is a dunce:

https://deadsimpletech.com/blog/altman_dunce