Real Ranker
Regular Member
- Jan 4, 2017
- 399
- 44
I Think Artificial Intelligence is far advanced and can beat Human Brain in no time... What do you think??
When do you think AI Robots gonna Rule the World??
- Thread starter Real Ranker
- Start date
then you shod study programming then you will understand how funny is that statement "oh no the finite state machine is going to take over the world".
or "the thing that saves sessions on database aka self learning is going to take over the world". mechanics have machine logic. to mimic human logic you need intuition (imagination).thats also why alot of people can understand algorithms but not implement them to a programming language, different logic type.
no computer can beat the human brain its the most superior machine you can get lost permanently in only a few parts of the brain. no one can recreate or understand how it works. humanity cant even replicate eye sensor tissue you want to create a human like machine? dream on. go watch some sci fi or fantsy movies.
you know how most ai pathfinding algorithms suck? chack out game ai and how much predictble stupid and buggy it is (exept for f.e.a.r bots they are cunning bastards with unit tactics) meaning the so called ai cant maneuver properly (cant walk) so how will it take over using facebook?
if your a troll your good, you shod apply for a trolland visa (its next to the douchebagistan embassy ).
sorry if im seen aggressive nothing personal i dont know you. its nerd rage.
or "the thing that saves sessions on database aka self learning is going to take over the world". mechanics have machine logic. to mimic human logic you need intuition (imagination).thats also why alot of people can understand algorithms but not implement them to a programming language, different logic type.
no computer can beat the human brain its the most superior machine you can get lost permanently in only a few parts of the brain. no one can recreate or understand how it works. humanity cant even replicate eye sensor tissue you want to create a human like machine? dream on. go watch some sci fi or fantsy movies.
you know how most ai pathfinding algorithms suck? chack out game ai and how much predictble stupid and buggy it is (exept for f.e.a.r bots they are cunning bastards with unit tactics) meaning the so called ai cant maneuver properly (cant walk) so how will it take over using facebook?
if your a troll your good, you shod apply for a trolland visa (its next to the douchebagistan embassy ).
sorry if im seen aggressive nothing personal i dont know you. its nerd rage.
Real Ranker
Regular Member
- Jan 4, 2017
- 399
- 44
Real Ranker
Regular Member
- Jan 4, 2017
- 399
- 44
It's not too far, I would say...
- Nov 7, 2009
- 2,833
- 1,427
This made my day! Thanks.
jazzc
Elite Member
- Jan 27, 2009
- 2,964
- 14,717
"Beat the human brain" doesn't mean anything. The oldest computer ever invented could easily make calculations faster than the brightest mathematician in the history. You wouldn't even use that computer today in an alarm clock. So the real thing is "on what" and to be more exact "on what that scares you".
RobertTheShark
Regular Member
- May 10, 2018
- 214
- 82
People are overestimating AI.
Reason for that is how the mainstream media is writing about it. AI has simply become a marketing word.
At the moment what mainstream media calls "AI" is just nothing else than better and smarter algorithms that solve problems that were already solved by "hardcoded" algorithms. It doesnt matte what we are talking about (text processing, image processing, voice processing etc.).
The reality is: we are far, far, far way from anything that is an artifical intelligence. The entire concept of todays "AI", wich basically is Machine Learning/Deep Learning is over 40 years old and was developed in the early 70s at the MIT computer labs.
The main reason we are seeing such a boom in Machine Learning is, because of the masses of data today available, organisations can trim their algorithms or "Machine Learning Apps" to be more precise about how they parse information. Thats the reason wyh AI was not really noticed in the last 40 years, today companys like Google or facebook simply have the masses of data to train their algorithms with them. If you have billions of datasets for images, you can make the detection pattern of your application way more precise, wich is what machine learning really is.
In the past I have done commercial image processing with - what you would probably call an AI - but in reality it was just a different approach with different algorithms. What you have to understand is, that a hardcoed algorithm can achieve a succes rate above 90% in parsing information, but with the concept of "Machine Learning" you can get like a few more percent out of the result. And thats, what in reality the big thing about "AI" at the moment is. Just getting like the last few percent out of it, making the detection rate for a car in a image (as an example) better.
Reason for that is how the mainstream media is writing about it. AI has simply become a marketing word.
At the moment what mainstream media calls "AI" is just nothing else than better and smarter algorithms that solve problems that were already solved by "hardcoded" algorithms. It doesnt matte what we are talking about (text processing, image processing, voice processing etc.).
The reality is: we are far, far, far way from anything that is an artifical intelligence. The entire concept of todays "AI", wich basically is Machine Learning/Deep Learning is over 40 years old and was developed in the early 70s at the MIT computer labs.
The main reason we are seeing such a boom in Machine Learning is, because of the masses of data today available, organisations can trim their algorithms or "Machine Learning Apps" to be more precise about how they parse information. Thats the reason wyh AI was not really noticed in the last 40 years, today companys like Google or facebook simply have the masses of data to train their algorithms with them. If you have billions of datasets for images, you can make the detection pattern of your application way more precise, wich is what machine learning really is.
In the past I have done commercial image processing with - what you would probably call an AI - but in reality it was just a different approach with different algorithms. What you have to understand is, that a hardcoed algorithm can achieve a succes rate above 90% in parsing information, but with the concept of "Machine Learning" you can get like a few more percent out of the result. And thats, what in reality the big thing about "AI" at the moment is. Just getting like the last few percent out of it, making the detection rate for a car in a image (as an example) better.
jazzc
Elite Member
- Jan 27, 2009
- 2,964
- 14,717
Worse yet, it's being brewed as a political battlefront. So this nonsensical fear-mongering is going to get worse.
Real Ranker
Regular Member
- Jan 4, 2017
- 399
- 44
slate_kane
Newbie
- Sep 29, 2016
- 18
- 11
HoNeYBiRD
Elite Member
- May 1, 2009
- 10,187
- 12,580
Not sure, if we will see that day, but when i read stuff like this: https://thenextweb.com/artificial-i...eloped-the-ability-to-navigate-like-a-mammal/, it really makes me wonder.
leadzepilln
Newbie
- Apr 19, 2018
- 6
- 3
When AI gains self-awareness like we humans do pretty much, some terminator type shit will happen.
ExperimentalSEO
Regular Member
- Mar 1, 2018
- 451
- 2,066
Much of the increases in for DL/ML is over-hyped by the tech companies and the media. Nvidia claims a 500x increase from 5 years ago, bycomparing the DGX-2 to a 580 SLI setup. Of course its a $400,000 machine vs $1500(?) machine.So its something like a 1.87x increase in performance per dollar, but a 25.85x increase in energy consumption per workload.
liamlio
Newbie
- Aug 28, 2017
- 46
- 22
Generally ai's today are self-optimizing algorithms that can improve when provided with mass amounts of data. They currently don't have the freedom to think or have any 'intelect', they are just well (sometimes not) optimized data.
On another note we haven't even reached the end of Moore's law, therefore it's unlikely that machine will have the capabilities to think within any of our generations. It would require to replicate what the human brain does, although an incredibly basic replication, however this is still potentially feasible in today's age/generation. However this would require an immense amount of computational power that just isn't economically affordable.
I personally believe as a nanotechnology engineering student that the future of computers and AI will lie in biological computers. Which are computers built using biological processes.
A good read to anyone's interested in this is this scientific paper where I professor create a basic computer using biology: http://rsif.royalsocietypublishing(dot)org/content/royinterface/14/128/20160990.full.pdf
Replaced the (dot) with a '.'; I can't post links yet
On another note we haven't even reached the end of Moore's law, therefore it's unlikely that machine will have the capabilities to think within any of our generations. It would require to replicate what the human brain does, although an incredibly basic replication, however this is still potentially feasible in today's age/generation. However this would require an immense amount of computational power that just isn't economically affordable.
I personally believe as a nanotechnology engineering student that the future of computers and AI will lie in biological computers. Which are computers built using biological processes.
A good read to anyone's interested in this is this scientific paper where I professor create a basic computer using biology: http://rsif.royalsocietypublishing(dot)org/content/royinterface/14/128/20160990.full.pdf
Replaced the (dot) with a '.'; I can't post links yet
ExperimentalSEO
Regular Member
- Mar 1, 2018
- 451
- 2,066
There is no such thing as vortex mathematics, its fake.