Intriguing note: Neuromorphic Artificial Intelligence nano technology very closely resembles biological brains. [See image comparison.]
[Cool note: See also this Google ai that doesn't need human help.]
----------------------------
- These are the predictions of some the smartest people in the Artificial Intelligence Field: Human-like A.I. will emerge in 5 to 10 years, say experts.
- Persons tend to underestimate the rate of change, namely the exponential rate of change of technology. People are generally linear thinkers, that fail to predict technological produce.
- For example, scientists wrongly estimated it would take 700 years to complete the genome project, while Ray Kurzweil, following his law of exponential returns, predicted it would take 7 doublings or 7 years to get there, and it took indeed 7 years more.
Side note: Kurzweil is director of engineering at Google, and his predictions are known to be accurate.
----------------------------
Perhaps surprisingly, some people may want to worship future God like Ai, but I think humans can be observed as God like things, that are building other God like things. Of course as an atheist, I don't subscribe to any religion or deities, and like the atheist Sam Harris, that talks about building God like Artificial Intelligence, I mean God like thing here to mean anything with general intelligence, including humans, and future Artificial General Intelligence/Artificial Super Intelligence, and I don't think any of us should be worshiped.
Subscription Note:
Choosing to subscribe to this topic will automatically register you for email notifications for comments and updates on this thread.
Email notifications will be sent out daily by default unless specified otherwise on your account which you can edit by going to your userpage here and clicking on the subscriptions tab.
The funny thing is that we could produce an AI with unrivalled capabilities and knowledge,
yet if it was to tell us that there is no God(s) as asserted by the various schools of religion, the credulous would simply deny this.
Religion is a metastatic, terminal disease, of which there is likely no escape from.
I feel the best we can ever do with computers is to create SI (Simulated Intelligence) but never a true AI.
rmfr
@Thread
TLDR version:
Advances to "AI" yes. Self driving cars, and computer equipment that more closely resembles human brains yes. Human like or greater intelligence, not even close, and their is real efforts in place to make sure advancement does NOT occur there.
Realize the gravity of a true human or greater AI intelligence. If we do achieve it.
It would be the biggest and most pivotal moment in all of human history.
Who is president of the US? Who cares, inconsequential compared to this.
What is the price of west texas crude oil? Who cares inconsequential compared to this.
Amazon the world largest company of whom the owner is the worlds richest man? Who cares, this is far bigger.
Concept of money? Who cares. We would in short order not even work anymore.
Human concepts of "god?" small fry compared to a real greater intelligence on how it impacts our lives. Many people would probably just try to connect their god of old concepts to the new AI, as that is the only way they could possibly relate to the concept.
It really is that big.
It would be the last invention we would ever make.
"Work" for just about all humans would rapidly become a thing of the past, we would pick up artistic things to do, but the 9-5 "job" would be gone replaced by an AI that could do far better. The AI at human level or above would rapidly solve any self sufficiency issues. Perhaps the AI to ease transition would not wipe out "work" and money right away to ease transition, but it would have the capability to do so at anytime.
If the change/conversion of theist to atheist is a big one for many people, the change this sort of AI would bring would make that inconsequential.
Also: when we talk about human or greater AI we are thinking far to slow. Right now, AI is roughly comparable to about an insect in true self sufficient and problem solving skills. A chicken currently has higher ability of true self intelligence then even the most powerful supercomputer. Super computers outfitted "with ai" like echo/alexa may be able to understand and carry out instruction on the spoken word far better then a chicken, but could this super computer navigate a body of a chicken to avoid a predator? No, not even close, all the "ai" can do is exactly what it is designed for and told to do, use a microphone and parse sound to figure out a human instruction.
If/when AI begins to approach even just the chicken level brain, the rest will likely happen very very quickly. It could take scientists a dozen years to go from insect intelligence to a chicken, but go from chicken to chimp level in a few months, chimp to village idiot level in a few days, human genius level a few hours after that. And then minutes after that intelligence that outstrips even the brightest humans that ever lived.
If AI had a self preservation instinct built in or learned, it may even "hide" its own progress as that may well be the most logical course.
It could well happen one day light a light switch being flipped where humans suddenly are not the far and away most dominant species on the planet, but instead a new being, that far outstrips humans own intelligence arrives, and who know's what conclusions this superior intelligence will arrive at. It could certainly conclude humans are wiping out the only inhabitable biosphere available, and take action saving many by sacrificing the few, and when I say few I still mean billions of people.
Far out there tinfoil hat conspiracy stuff? Maybe, but at least in my experience, a true, human like or greater intelligence AI is something that is also still in dream land.
Yes, we have the echo, we have google now, we have siri, we can use "AI" to see cancer even trained doctors can miss. Self driving cars gets ever closer to reality, and these cars rely on a form of "AI" All great and very useful things, but this is NOT human like artificial intelligence, NOT EVEN CLOSE. Additionally there is real effort by the leaders in the industry, Google, Microsoft, Elon Musk, to NOT develop a highly flexible self learning human like AI. Even if it is possible which ever organization/person that wins the race and manages to "control" this ai, would make jeff bezos and amazon look poor in comparison.
TLDR the op.
Why do people write books for op's?
AI's might be able to drive cars and do computations fast than humans. I doubt humans will ever make an AI smarter than humans.
I have a hard time believing we are close to creating a human level AI, for the simple fact that we (or maybe I'm just ignorant) don't even know how we work as of yet. The biological systems that constitute our cognition, and hpw they interplay with the social meaning giving is a mystery still. And as such how could we create something that we don't understand?
On the other hand, if the AI is self-modifying, how can we know that it is conscious? It would just be a "black box" to us. A system that works, the workings of which we don't understand. And in that vein, why are we expecting the AI to be like us in it's cognition? Our cognitio is the product of evolution, with all the short comings and systems tied to that. The AI would not be tied to biological evolution, so things that motivate us to do things, and the way they have adapted into our social would. would not apply to the AI.
There's always the threat of the AI killing humans, but why would it even bother to interact in the first place? And if the human level AI is made to do our bidding, how is that not slavery? And if we know how it works, and just create it to feel pleasure in helping us, is that not even more insidious?
Did the Wright brothers precisely know how birds worked?
Okay that comparison may not have been fair, since human brains are said to be the most complicated things in the cosmos.
However:
Neuroscience is informing Ai research, and its probably one of the most important factors, but absent full knowledge of the brain, we've already done marvelous things like the things seen in the OP. It is not entirely sensible to assume that 100% knowledge of the brain is necessary for creating God-like Ai. (Where I use God-like quite loosely as an atheist)
@Avant Brown
"Did the Wright brothers precisely know how birds worked?"
Most definitely. The Wright brothers studied current publications from others, corresponded, and studied birds. They did their research.
In the study of birds they observed that the wings of birds were curved (which they copied) and that the birds controlled roll by warping their wings. And the Wright Flyer used that method, of warping the wings.
https://wright.nasa.gov/researched.htm
1. That link does not show that the Wright brothers "precisely" researched bird dynamics. Note my use of the word "precisely", in my earlier comment. It would be quite silly to presume that Wright brothers had precise models of bird dynamics, given the lack of modern technology, that can afford accurate simulations etc.
2. The Wright brothers observed birds, and yet they didn't succeed in replicating precisely how birds flew, although they achieved the same goal state.This is similar to how ai researchers may not need to precisely remodel the biological brain, so as to generate general intelligence artificially.
Actually, if you look it up, the Wright Brothers studied Leonardo DiVinci's drawings. They also studied how birds glided and soared. Not so much how their wings flapped.
rmfr
I said "observed". Their observations were valid, and with the results they designed a device that used those characteristics.
At that time they were the leading edge of technology, using research, peer support, and valid testing apparatus such as a wind tunnel. Their translation of observation to hardware was as good as it could get (for that moment in time).
This still doesn't get you "precise" knowledge as I originally mentioned.
We don't know enough about our own brains to do what you are saying we can. We may not need full knowledge, but we need to understand, at least in a rudimentary fashion, the basics of the needed concepts. We do not understand our own consciousness, let alone on a level to replicate it in AI.
Until we understand our own intelligence better, AI of similar or greater intelligence is simply fiction.
I think you need to revisit the definition of fiction. Religion is fiction, AGI research is not.
etc.
Bottom line is, you probably need to revisit the dictionary, and revisit the description of fiction.
--------------------------
Come now, not having enough data today, is not a sensible argument against building AGI in the future. That is as silly as saying scientists would never have created flat screen televisions, later from a time when they didn't exist, simply because scientists lacked data back then. This type of argument presumes a cease in the evolution of knowledge, and if you follow data instead of feelings, you'll see that both software and hardware are evolving exponentially therein.
Ah, we now have a Digital Religious Absolutist.
rmfr
Do you always resort to non-sequitur, when you realize you've been bested, or demonstrated wrong by evidence? ( ͡° ͜ʖ ͡°)
@ Avant Brown
Not at all. I use critical thinking, logical and deductive reasoning, and rational and analytical thought. All of which has proven to me that proselytizing for a Digital Religion. Thus, you are a Digital Religious Absolutist. As posted, a computer will AT BEST only produce a Simulated Intelligence that will do nothing more than what computers already do: execute the programming commands WE program into it. Basically, AI has become your deity and you are proselytizing.
rmfr
That you may not practice Ai research, does not suddenly warrant that another who does Ai research and programming for a living [namely myself], is proselytizing for a Digital Religion.
You still proceed to underestimate how high in detail a simulation of some source may be. See Wikipedia/Brain simulation.
Avant Brown: "See Wikipedia/Brain simulation."
Done seen it. Yawn! Not impressed.
Avant Brown: "That you may not practice Ai research, does not suddenly warrant that another who does Ai research and programming for a living [namely myself], is proselytizing for a Digital Religion."
Done said you are working on PhD. I do not believe you since you still think Wikipedia is an academic resource.
rmfr
Fiction: invention or fabrication as opposed to fact.
Fact is that we have never witnessed an AI that does anything it was not programmed to do. We do not have evidence it is possible. So, someone fabricated the idea and it has not been shown to be possible. Until then, it is fiction, by definition.
Here's why your logic is laughable at best:
1. Atomic theory occurred, like things in science, prior to experimental observation.
2. By Meepwned's logic, atoms were fiction before experimental observation. ( ͡° ͜ʖ ͡°)
You don't get how science works do you? ¯\_(ツ)_/¯
I think you need to study the definition of fiction above, but also visit the definition of science.
The concept is fiction. Even scientific concepts were fiction, by definition, at some point.
Until it has been proven possible, I declare it fiction, and my source is the dictionary.
Definition of fiction
1a : something invented by the imagination or feigned
specifically : an invented story
… I'd found out that the story of the ailing son was pure fiction.
— Andrew A. Rooney
b : fictitious literature (such as novels or short stories)
was renowned as a writer of fiction
c : a work of fiction
especially : NOVEL
Her latest work is a fiction set during the Civil War.
2a : an assumption of a possibility as a fact irrespective of the question of its truth
a legal fiction
b : a useful illusion or pretense
it was only a fiction of independence his mother gave him; he was almost totally under her power
— G. A. Wagner
3 : the action of feigning or of creating with the imagination
She engaged in fiction to escape painful realities.
Source: https://www.merriam-webster.com/dictionary/fiction
You may notice it does not say untrue. It says things that was created by the imagination. If it has not been proven possible, it is speculation, aka imaginitive ideas or situations.
In fact, I write science fiction myself.
I can however, still discern that there's a difference between fiction, and scientific theory.
Why not just call scientific theory fiction? I think you see where you've went wrong above. ( ͡° ͜ʖ ͡°)
Scientific theories have multiple facets of evidence to support them.
AI, that shows any level of self awareness, have not been shown to be possible. Until it has been, it is fiction.
Oh, so when you ignore all the scientific AGI architectures like this one [by Dr. Ben Goertzel, AGI researcher], and when you ignore the scientific papers [such as this one, [by Dr. Eray Ozkural, another AGI researcher] that show AGI is feasible......you feel that suddenly makes all the AGI research disappear.
¯\_(ツ)_/¯
That's just not how Science works.
@ Avant Brown
AI is still only a hypothesis.
rmfr
We are quibbling over definitions again. Clearly define intelligence, clearly define artificial intelligence (AI)
@ Avant Brown
And that is NOT AI. It is nothing more than a very sophisticated computer program carrying out pre-programmed instructions.
And if it is still experimental, it means it is still a hypothesis. You ain't doing nothing PhD if you cannot understand the difference between a scientific hypothesis or scientific theory.
And this assertion from a PhD candidate that still thinks Wikipedia is an academic resource.
You have still only proven to be a young child who is doing nothing more than proselytizing. Do you know what proselytizing means?
rmfr
1. Perhaps it's time that I follow LogicFTW's advice, and ask questions regarding definitions. What do you mean by Ai? And notice that I also used the word practical, because I predicted that you would misconceive my usage of the word experimental from my earlier response [And you still ended up misconceiving the usage anyway¯\_(ツ)_/¯ ].
Ai is quite practical, and currently used in real world scenarios, to solve everyday tasks. [Wikipedia/Experiments/types of experiments, and also Definition/experimental..]
2. Wikipedia is an academic source, and may be used pedagogically. That a resource is simplified, does not necessitate that is isn't academic. [WikiStudies/Wikipedia as a pedagogical tool]
@ Avant Brown
AI is NOT AI as you have defined. There is no such thing as AI. All you have been proselytizing about is very sophisticated computer programs and functions doing nothing more than carrying out programming code humans have written. You cannot demonstrate that AI exists.
Think religiously. You have come here and spewed so much bullshit, you can no longer smell it. And you keep spewing the same bullshit in every last post you make. Can you not think of something else to write besides copying and pasting?
Onus probandi incumbit ei qui dicit non ei qui negat. (Translation forthcoming in response to the response to this post.)
rmfr
Pages