Study Suggests Artificial Intelligence GDP Impact Comparable to Steam Engine

cageymaru

Fully [H]
Joined
Apr 10, 2003
Messages
22,090
McKinsey Global Institute has released a new study that compares the potential impact of artificial intelligence (A.I.) to the invention of the steam engine in driving economic growth for the countries that are first to establish themselves as leaders in the field. It is estimated that A.I. could add 1.2% to the annual gross domestic product (GDP) for at least a decade. 70% of companies will adopt a form of A.I. by 2030 and it is expected to add $13 trillion in economic activity or 16% higher cumulative GDP compared to today. Countries that are first can expect to capture an additional 20 to 25% in net economic benefits thus widening the digital divide between them and the rest of the world. Nonadopters could see a shrinkage of 20% in cash flow.

AI uses large data sets and algorithms to mimic human behavior. The world's two largest economies, the U.S. and China, are both racing to invest heavily in the technology. Beijing, in particular, has made AI part of its five-year plan that runs through 2020 and wants to become a leader in the technology by 2030, the McKinsey report pointed out.
 
The categories discussed...

computer vision
natural language
virtual assistants
robotic process automation
advanced machine learning

While a true AI would be capable of any of those, even when combined those don't make for an AI.
 
AI is not a revolutionary idea. It is an evolution in computer programming.

and they must have forgotten to add that while GDP is expected to continue to grow, real wages and purchasing power will continue to deteriorate , except for countries whose economies are in dire straits now, due to a lower benchmark

wake me up when AI makes purchasing parity back to 50 years ago...
 
Dude ibm is just killing it with watson, i mean its like printing money. They are not laying off anyone in their AI section, and are rigorously hiring.
 
"70% of companies will be using AI by 2030..."

Why do i get the feeling that their definition of A.I. and my definition are completely different. I would be shocked if A.I. actually exists by 2030...

Isn't the proper name for what is being peddled as "AI" actually "Simulated Intelligence"
 
"70% of companies will be using AI by 2030..."

Why do i get the feeling that their definition of A.I. and my definition are completely different. I would be shocked if A.I. actually exists by 2030...

Isn't the proper name for what is being peddled as "AI" actually "Simulated Intelligence"

You really want to get hung up on their use of "AI" instead of ANN. AI is an easier term for marketing people to spin... average folk don't get excited that their new widget/service is using an Artificial neural network. We all know what they are talking about. No the androids are not coming... but 100% of systems are already using some form of ANN. Anyone that uses google search, facebook and pretty much any other online web service at this point is benefiting from computer systems employing ANN systems.

Going forward their use is going to revolutionize everything framing, manufacturing, medicine... you name it even gaming if Nvidia is to be believed.
 
Last edited:
You really want to get hung up on their use of "AI" instead of ANN. AI is an easier term for marketing people to spin... average folk don't get excited that their new widget/service is using an Artificial neural network. We all know what they are talking about. No the androids are not coming... but 100% of systems are already using some form of ANN. Anyone that uses google search, facebook and pretty much any other online web service at this point is benefiting from computer systems employing ANN systems.

Going forward their use is going to revolutionize everything framing, manufacturing, medicine... you name it even gaming if Nvidia is to be believed.

Not trying to get caught up on anything, just that when I hear AI, probably like 90% of people I think of a computer that has gained self awareness instead of a bunch of algorithms that can mimic self learning. I just don't see any difference in the increased application of computing power that has been driving the information age already and suddenly throwing some algorithms that keep and draw from statistical data sets to help make decisions.

I mean, lets not get ahead of ourselves here, that's really all it is, it's just that processing power has finally gotten to the point where it can collect enough data points via brute force and slapping a Boolean spool on it.
 
Last edited:
The categories discussed...

computer vision
natural language
virtual assistants
robotic process automation
advanced machine learning

While a true AI would be capable of any of those, even when combined those don't make for an AI.

Alright. So what would it take for an entity to qualify as an AI?
 
Not trying to get caught up on anything, just that when I hear AI, probably like 90% of people I think of a computer that has gained self awareness instead of a bunch of algorithms that can mimic self learning. I just don't see any difference in the increased application of computing power that has been driving the information age already and suddenly throwing some algorithms that keep and draw from statistical data sets to help make decisions.

I mean, lets not get ahead of ourselves here, that's really all it is, it's just that processing power has finally gotten to the point where it can collect enough data points via brute force and slapping a Boolean spool on it.

That is not at all in fact what is happening. Its not brute forcing anything. "AI" ANN in fact takes less processing power there is no brrute force and boolean search methods going on.

A standard computer has a central processor that can address memory locations where data is stored. The processor reads data the instruction requires from memory, the instruction is then executed and the results are saved in memory. In a traditional system the computational steps are deterministic, sequential and logical, and the state of a given variable can be tracked from one operation to another.

ANNs are not sequential or necessarily deterministic. There are no complex central processors, rather there are many simple ones which generally do nothing more than take the weighted sum of their inputs from sensors/other processors whatever data set. ANNs do not execute instructions; they respond in parallel to the pattern of inputs presented to it. There are also no separate memory addresses for storing data. Instead, information is contained in the overall activation 'state' of the network. 'Knowledge' is thus represented by the network itself, which is quite literally more than the sum of its individual components.

Its a completely different way of doing computing... and there have been people pushing the idea since the 60s, and most people assumed they where completely insane. It hasn't been until the last handful of years or so that we have been able to create enough connections that move fast enough to have it actually work.
 
Alright. So what would it take for an entity to qualify as an AI?
Dynamic adaptation to unknown and unrelated tasks to begin with.

Everything called AI in news nowadays are kind of sophisticated statistical tools. There is more intelligence in a guppy even if it can't play go than in anything made by humankind combined.

There is something subtle to the autonomous processes of the brain that when combined together creates intelligence, that today is still beyond our knowledge.
 
Dynamic adaptation to unknown and unrelated tasks to begin with.

Everything called AI in news nowadays are kind of sophisticated statistical tools. There is more intelligence in a guppy even if it can't play go than in anything made by humankind combined.

There is something subtle to the autonomous processes of the brain that when combined together creates intelligence, that today is still beyond our knowledge.

See, you can't define intelligence. Sure we don't have AI yet... I think. But on the internet you never know. I could be an AI. Have you lost any games online recently? Probably AI. Better yet alien AI.
 
You:
That is not at all in fact what is happening. Its not brute forcing anything.

Also you same post:
It hasn't been until the last handful of years or so that we have been able to create enough connections that move fast enough to have it actually work.

When you are parsing several million data points and cross referencing it with several million data points to determine what "Y" and "Z" are most likely will be in a standard boolean spool of If "X" then "Y" else "Z" it takes a lot of brute force processing power.

I'm not trying to say you are wrong, hell I may well be an idiot, but that's all I'm seeing any of this to be is using the exact same methodology used by scientific prediction models that have been around forever and applying that to the business sector now that processing power has made it financially viable to utilize.
 
See, you can't define intelligence. Sure we don't have AI yet... I think. But on the internet you never know. I could be an AI. Have you lost any games online recently? Probably AI. Better yet alien AI.

Sure you can. Take any "AI" and throw it in a completely foreign scenario unrelated to it's purpose and if it can adapt to that scenario on it's own and do that consistently without any input, than it's intelligence.

Example: Take the Google search engine "AI" and without making any changes to the code apply it to self driving cars, then apply it to the burger flipping robot, and see how good it is at playing League of Legends or whatever. If you can consistently throw it on any task and it improves, it has intelligence.
 
See, you can't define intelligence. Sure we don't have AI yet... I think. But on the internet you never know. I could be an AI. Have you lost any games online recently? Probably AI. Better yet alien AI.
Well I also can't define blue, but when something isn't blue I know it.

A recurrent nightmare for AI researchers is achieving AI -the real deal- somehow and having no clue of how that happened.
 
well, intelligence is about making the right decision when given a variety of input and choices. And as with pretty much every intelligent being in the universe, they can be intelligent in one task but extremely stupid in another. Or just outright impossible to attempt. So technically , what we have is AI now.

what we conceptualize AI in the scifi books is to be is something that is also imbued with limited sentience and creativity. So far, scientists and psychologists do not even dare say they know how a dog's brain work , not to mention human brains. So whatever AI we will have in the near future will be nothing more than coded simulations, albeit a realistic enough one to pass off as genuine enough in short simple interactions.
 
Back
Top