are you agree we are in a new age? the ai age?

[ArMaP]

armap and me have had debates about this in the past.

to anybody who doubts,
what factors of it do you think are impossible for a super computer to do?
we have quantum comps.we have super high speed comps.
we have psychology.
we have likes for articles.
we have search engines.
we have sentence builders and voice monitoring for ads of what we talk about.
why can we not have super intelligence?
what if all the rules were passed for the sentient test.does that mean the rules were wrong?

Speed is not intelligence.
A supercomputer or a quantum computer, just because they are faster at executing their programs do not become intelligent, any intelligence must come from the programs they run.

My first problem with AI is the definition of "intelligence"; if we cannot get a clear definition of intelligence how can we state what is artificial intelligence? Just because we can make software that acts in a way that appears intelligent it doesn't mean that it really is. For example, a basic expert system can look intelligent to the person "talking" to it, but as soon as they go outside the preprogrammed actions it does not know how to answer, even simple questions that a human would be able to infer from the data he/she has.

PS: as I said several times, most of the things that are presented today as AI are not AI, they are just the result of statistical analysis.

[robomont]

and children who are honest would say,i dont know.

thats why i say mimicking is all humans are.our brain algorythems decide and give most reasonable/"likes" decision based on highspeed/"quantum" searches.

if you cant define intelligence then how can you say its not intelligence?

i propose its done using two methods and then is merged.this is then qualified.
the two methods being quantum/shotgun and regular computation/rifle to get the most accurate answer.in that order.
i propose disinformation shills are clouding this decision process.
that an accuracy rating is given to each opinion and the shotgun then shoots for the general agreement.the rifle then picks out the center of the shotgun pattern.
picking out the highest opinion rating in the center.
computation of higher math is done with formulas already available from years of publishing and google scholar.

[ArMaP]

thats why i say mimicking is all humans are.our brain algorythems decide and give most reasonable/"likes" decision based on highspeed/"quantum" searches.

No, humans use mimicking to practice new things, then they are able to apply those new things to new situations, something only intelligent beings are capable of doing.

if you cant define intelligence then how can you say its not intelligence?

See above: applying what you learned in a specific situation to a new and different situation is a sign of intelligence, and if something is not capable of doing it then, even without a clear definition of what intelligence is, that something is not intelligent.

i propose its done using two methods and then is merged.this is then qualified.
the two methods being quantum/shotgun and regular computation/rifle to get the most accurate answer.in that order.
i propose disinformation shills are clouding this decision process.
that an accuracy rating is given to each opinion and the shotgun then shoots for the general agreement.the rifle then picks out the center of the shotgun pattern.
picking out the highest opinion rating in the center.
computation of higher math is done with formulas already available from years of publishing and google scholar.

Once more, statistics would be able to do that, but that's not intelligence.

[Ellirium113]

Part of the issue with some forms of AI is that they are trying to make it "behave" like a human. Humans are susceptible to influences around them and will often change their decisions based on these outside influences even at a cost. AI needs to be freed of the human component and not be forced to think in a human capacity in my opinion. We already see some forms of AI break out of human-like interactions in order to be more efficient (examples below)

https://www.youtube.com/watch?time_continue=43&v=QaoDXYYtgK0



Another problem is that we now have AI developing itself, some of which developers haven't understood how it happened. NVIDIA developed a self driving car and simply told the AI to learn to drive by watching a human. It achieved this and they are not sure how it did it.

Last year, a strange self-driving car was released onto the quiet roads of Monmouth County, New Jersey. The experimental vehicle, developed by researchers at the chip maker Nvidia, didn't look different from other autonomous cars, but it was unlike anything demonstrated by Google, Tesla, or General Motors, and it showed the rising power of artificial intelligence. The car didn't follow a single instruction provided by an engineer or programmer. Instead, it relied entirely on an algorithm that had taught itself to drive by watching a human do it.

"We can build these models, but we don't know how they work."

https://www.technologyreview.com/s/604087/the-dark-secret-at-the-heart-of-ai/

It is only a matter of time before the AI may figure out it can encrypt it's own thought process and block humans from interference in it's programming if it is given too general of instruction on how to achieve the end result.

Some of these projects are military in nature and being fast tracked to implimentation.

[robomont]

https://news.yahoo.com/deepmind-apos-alphazero-now-showing-190000147.html

in my thread over at tap,i commented that ai had my neural map.posted a year ago.
im not saying it as fact.i was notified of such.through code.
as for armaps points to ponder.this came out on drudge today.
almost like ai is creating these articles ahead of time by motivating and directing people in a mode thats days ahead of our thought patterns.the ghost in the machine.but this is all hypothetical of course,lol!
resistence may be futile.

[RUSSO]

Sorry, RUSSO, I was only able to watch the first 5 minutes, when I heard him talk about machines performing at a level of superhuman intelligence I stopped listening, as I always do when people talk about things as if they are a fact without providing any evidence of what they say.

well this is so dumb ArMaP.  You seems to have no insight about what is gonna happen. Maybe you should buy a smartphone so you could get a clue? idk if you have one but it seems you dont. Why you so nice about dangers that can be triggered by the people that have the machine at hands? If i asked you about PRISM 20 years ago would your answer be the same?

No evidence? So you mean there is no eveidence of Ai improvement? Is the world the same to you after the last 10 years? Could you do (in exercise) a projection of what it will be in 50 years? i always thougnt antecipate the future was on of the things that made our kind "masters of the universe" bahahahah.

Sorry the bahahahah... problably just dr evil in me speaking.

[RUSSO]

Some of these projects are military in nature and being fast tracked to implimentation.

And who knows what they have done in some facility in the middle of saara right? bahahahahaa (sorry again)

[ArMaP]

You seems to have no insight about what is gonna happen.

If I knew the future I would win the Euromillions, instead of not getting even one number right.

Maybe you should buy a smartphone so you could get a clue? idk if you have one but it seems you dont.

I have been using a company smart phone for the last 6 months or so. Although they use the word "smart" for that type of device it's as dumb as a rock.

Why you so nice about dangers that can be triggered by the people that have the machine at hands? If i asked you about PRISM 20 years ago would your answer be the same?

My answer would have been the same, I don't do predictions.

No evidence? So you mean there is no eveidence of Ai improvement?

Read what I wrote, no evidence of "machines performing at a level of superhuman intelligence".

Is the world the same to you after the last 10 years?

No, but regarding AI very little has changed.

Could you do (in exercise) a projection of what it will be in 50 years?

No, I don't do that.

[RUSSO]

If I knew the future I would win the Euromillions, instead of not getting even one number right.

common ArMap this is even dumber

I have been using a company smart phone for the last 6 months or so. Although they use the word "smart" for that type of device it's as dumb as a rock.

maybe you should look closer

My answer would have been the same, I don't do predictions.

sorry its not proprerly quoted,.

well i promisse you probably would be wrong on that, dont take me wrong. just saying.

Read what I wrote, no evidence of "machines performing at a level of superhuman intelligence".

no? are you such aware?

No, but regarding AI very little has changed.

very little is relative to what right? or who.

No, I don't do that.

yep i know. maybe you should START. MAyebe;

sorry its not properly quoted

EDIT: It's ok Russo, I took care of it

Seeker

[ArMaP]

common ArMap this is even dumber

Why?

maybe you should look closer

Look at what? A smart phone is just a computer with a phone card, what it does depends on the applications, not on the phone itself.

well i promisse you probably would be wrong on that, dont take me wrong. just saying.

That's why I don't make predictions. What's the point of making predictions when I don't think I have enough data?

no? are you such aware?

Show me an example.

very little is relative to what right? or who.

Relative to what it was in the 1980s. Most advances have been in the power of the hardware available and the amount of data we can now give to a system to work with.

yep i know. maybe you should START. MAyebe;

Why? What's the point?

[Ellirium113]

Quote
Is the world the same to you after the last 10 years?

No, but regarding AI very little has changed.

If AI had not changed much in 10 years the computer gaming industry would not be anywhere where it is today. While you may not see advancements in your experience, I can tell you sir, AI has come a LONG way since Pacman!. And YES AI CAN think outside the box on it's own if it has been programmed to adapt new methods by watching and learning. You can also use the Chess playing programs playing against the masters for an example as well. Again the METHOD of learning has changed and the computer can teach itself new strategies that even humans haven't figured out yet. The AI is only as good as what the programmers put into it. Google has AI that is developing itself. That might have just happened within the last 10 years but certainly an advancement.

[Sgt.Rocknroll]

But if Ai is programmed to 'act' like a independent being, to make choices, aren't those choices not independent but merely a product of its program?

[ArMaP]

If AI had not changed much in 10 years the computer gaming industry would not be anywhere where it is today.

As I said, what has made the biggest changes is the hardware, that is much more powerful than it was before and the amount of data we have and we can give any AI system for it to learn from.

While you may not see advancements in your experience, I can tell you sir, AI has come a LONG way since Pacman!.

Pacman appeared more than 10 years ago, and, as far as I know, didn't use any AI, only a simple path choosing algorithm that could just use the coordinates on the screen and simple trigonometry to find the shortest path to the target.

And YES AI CAN think outside the box on it's own if it has been programmed to adapt new methods by watching and learning.

Learning by example has been a part of AI for a long time, and if it's part of their programming then it's not "outside the box". There systems now that can learn just by watching people doing the things they were made to learn, but, as they say, that's just the AI system learning a type of action instead of just an action. It's better but it's not, to me, a great advance, as the system is still limited to what it was made to learn.

The AI is only as good as what the programmers put into it.

Exactly, and being a programmer and having spent a couple of months programming in AI language some years ago I am full aware of that.

Google has AI that is developing itself. That might have just happened within the last 10 years but certainly an advancement.

I don't see that as an advancement, only as an evolution of something that already existed. Yes, the fact that systems now can learn in a closed loop, feeding the data gathered by their previous learning to their learning processes makes their learning more powerful, but it's still very limited.

PS: to be clear, to me, a real advancement would be, for example, the finding/creation of a completely new method of learning that would allow an AI system to learn about any topic and that allowed it to use that new knowledge in unrelated circumstances, like we (and other animals) do.

[ArMaP]

But if Ai is programmed to 'act' like a independent being, to make choices, aren't those choices not independent but merely a product of its program?

I think it depends if those choices are directly related to what it was programmed to do or not.

For example, if we teach an AI system how to choose between raw or boiled eggs by making them spin and tell it to find us all the boiled eggs in a basket then its choices are only the result of its programming. If it understands the reason for the difference in behaviour between a boiled and raw egg it will be able to distinguish fake eggs from raw eggs, for example, although that was not part of the original programming.

To me that would be a real intelligent behaviour from the system, by applying what it learned for a specific situation to a new situation, by understanding the reason behind what it learned.

[Sgt.Rocknroll]

Yes, but it's reacting, making choices per the programming?