Humans and Artificial Intelligence

AI (Artificial Intelligence) is coming - it's already showing up in our consumer products, and it's going to continue to increasingly make itself felt in our economy. It is reducing the difficulty of doing certain jobs, and even taking over and replacing certain simple job roles that were being done by humans. There are some things machines can do better than humans - particularly repetitive tasks which require a lot of precision.

Should AI be regulated? If so, then in what ways?

As AI continues to take over and eliminate jobs from the economy, what should be done to address the unemployment that will grow from this?

Finally, should we be concerned that AI will ultimately pose a threat to humanity itself? What can we do to avoid this, while reaping the benefits of AI?

Merging AI with Human Intelligence - Should We Really Go That Far?

sanman said:

AI (Artificial Intelligence) is coming - it's already showing up in our consumer products, and it's going to continue to increasingly make itself felt in our economy. It is reducing the difficulty of doing certain jobs, and even taking over and replacing certain simple job roles that were being done by humans. There are some things machines can do better than humans - particularly repetitive tasks which require a lot of precision.

Should AI be regulated? If so, then in what ways?

As AI continues to take over and eliminate jobs from the economy, what should be done to address the unemployment that will grow from this?

Finally, should we be concerned that AI will ultimately pose a threat to humanity itself? What can we do to avoid this, while reaping the benefits of AI?

Merging AI with Human Intelligence - Should We Really Go That Far?

Click to expand...

AI =/= automation.

sookster said:

AI =/= automation.

Click to expand...

But AI is becoming part of automation - the technology is continuing to evolve and getting more and more powerful. It's not like we're living in a world/economy without AI, and one fine day we'll wake up in a world with AI, after some switch has been flipped. AI is gradually emerging, gradually becoming more capable and more powerful. And it's going to continually and increasingly affect us and how we live.

sookster said:

AI =/= automation.

Click to expand...

I don't think he ever claimed they were the same, but they are unequivocally working together to replace human jobs. We can see this as a big tragedy or we can see it as a relief that we have less menial jobs and more people can focus on work they actually care about. That all depends on how we choose to structure our relationship to work. Should our very survival be linked to our ability to find jobs in an economy with fewer and fewer jobs yet more and more people?

Tim the plumber said:

Today it is a little like asking what regulations should the aviation industry use in 1890.

When we have something that is actually intelligent we should work out what the regs should be.

Click to expand...

The Write brothers didn't fly until 1903. We are way past the point of the most rudimentary example of the technology. We already have low level AI and it's growing exponentially. More and more jobs are being replaced as the capabilities of automation and AI grow. The time to start thinking about solutions and regulations is now, not 20 years into the problem.

We don't need to have immaculate AI to replace menial American jobs, but when we do, there will be very little need for the labor that the vast majority of humans can provide.

I think one of the things AI has in its favor, is that it will help to lower the cost of living - after all, nobody will adopt a new technology that raises costs - technology adoption is driven by cost savings. So AI should bring down the costs of goods and services.

While AI will eliminate jobs, it will also create some new ones, some relating to design/operation/maintenance of AI-enabled products themselves.

But there will also be existing industrial sectors that are freshly empowered by AI, just as there have been many traditional industrial sectors that have been empowered by other technologies, and these sectors will be able to expand their scope due to the availability of AI as a new technological enabler.

What industrial sectors are most likely to benefit from AI?

sanman said:

I think one of the things AI has in its favor, is that it will help to lower the cost of living - after all, nobody will adopt a new technology that raises costs - technology adoption is driven by cost savings. So AI should bring down the costs of goods and services.

While AI will eliminate jobs, it will also create some new ones, some relating to design/operation/maintenance of AI-enabled products themselves.

But there will also be existing industrial sectors that are freshly empowered by AI, just as there have been many traditional industrial sectors that have been empowered by other technologies, and these sectors will be able to expand their scope due to the availability of AI as a new technological enabler.

What industrial sectors are most likely to benefit from AI?

Click to expand...

Hmm... If the cost of X, say a home, goes down, by say 20%, from AI (robots?) replacing human labor then how does that benefit those left unemployed to achieve that cost savings? Will we then pay non-workers the same as workers? More importantly, how will we determine who among us is to be which?

If Skynet is willing to give me a house by the beach, beautiful women and free food I shall worship the machine.

ttwtt78640 said:

Hmm... If the cost of X, say a home, goes down, by say 20%, from AI (robots?) replacing human labor then how does that benefit those left unemployed to achieve that cost savings? Will we then pay non-workers the same as workers? More importantly, how will we determine who among us is to be which?

Click to expand...

The market tends to seek an equilibrium, whereby the desired prices of both buyer and seller intersect. So if houses become unaffordable and unbought because of the aforementioned unemployment, then their price will automatically go down. Otherwise, what's a homebuilder going to do - build a home using his robot, and just sit on it?

What will happen during all this buyer-seller price-seeking, is a change in the way the robotics technology is applied/implemented/used.
That initial high-tech homebuilder you mentioned will only achieve the 20% reduction in the cost of construction, and won't be able to get adequate demand for the homes he gets built.
Somebody else will change the way the robotics/AI is used in the homebuilding, so that he achieves 40% reduction in the cost of construction. Somebody else will achieve 60% reduction in the cost of construction. Etc, etc.

But for these changes to happen, the robotics/AI has to at least start getting used in this sector.



The above is merely an example in automation - I only posted it because you mentioned homebuilding - but Artificial Intelligence is more about giving machines the smarts to deal with the complexities of the real world. There are related component technologies, like machine vision, optical recognition, voice recognition, LIDAR, etc, which give AI the tools it needs to grapple with the unconstrained real world.

The driverless car is an example of a machine performing a complex task that would normally require humans to perform. But of course, a driverless car can work 24/7 - it doesn't get tired, or sleepy, its reflexes don't degrade, its judgement can be continually improved just by learning from past data, and processing hardware will continue to improve each year, unlike the human brain, so that performance improves.

Stuff like machine vision may have started out being used in more controlled places like factory environments (eg. sorting parts or spotting defective ones on a conveyor belt), but obviously they're now increasingly moving out into the uncontrolled real world.

Perhaps battlefield applications will be the real stress-test that will force AI to evolve much higher levels of performance, so that it will be much more ready for ordinary everyday life. On the other hand, the more recent trends have seen consumer-level computing technologies advancing faster than military ones, because consumer applications are far more widespread and active than military ones.

But the more the technology advances, the more cases it will be open to use for, thus increasing the spread/scope of its usage, thus giving more opportunities to evolve/improve, including via machine learning.

This is how we get plugged into the Matrix

You have to rid of society of capitalism when the robotic revolution hits full gear.

https://www.thevenusproject.com/

AI is meant for the purpose of interacting with the real world - the complicated, unfiltered, uncontrolled real world that we humans have evolved to function in.

AI has certain advantages, like hardware components which keep getting better and better every year. Or the fact that it doesn't get tired like the human brain does. The AI truck driver or AI airline pilot can work across the long-haul trips, and not need any rest breaks. AI doesn't draw a salary, so its cost is in R&D and maintenance.

But of course the obvious drawback is that AI isn't born into this world like a human, being raised like a human, learning and being taught like a human is, gaining all that valuable common sense about human context in the process.

So AI has to start out in controlled environments in the beginning, and perhaps widen its scope of capability from there.

Imagine being the AI ditch-digging machine. You're introduced to the world living in a ditch. You dig and dig, and that's your world. You learn or are taught stuff along the way, like if you hit a pipe and liquid starts squirting out, you'd better stop digging right away. And that's your world - maybe that's all you need to know.

Or what about an AI cab? You'd better come with a lot of smarts off the bat, so that you don't run over people, know how to handle traffic and avoid collisions, and can deal with occasional detours due to road work - all while getting your passengers to their destination in reasonable time.

When it comes to machine learning, the more that machine interacts with the world, the better it can get at handling it to achieve the desired outcome. The learned knowledge can of course be extracted and shared with all the other machines, for an incredibly powerful collective learning.

The most interesting learning and interactions may happen when we have machines that can conveniently go anywhere under their own control, and experience all the situations that humans can experience, and more. That will enable machines and their collective to learn all the intricacies of the human experience, to know all that we know. That is perhaps the romantic dream of a robot which is our creation in the image of ourselves.

sanman said:

The market tends to seek an equilibrium, whereby the desired prices of both buyer and seller intersect. So if houses become unaffordable and unbought because of the aforementioned unemployment, then their price will automatically go down. Otherwise, what's a homebuilder going to do - build a home using his robot, and just sit on it?

What will happen during all this buyer-seller price-seeking, is a change in the way the robotics technology is applied/implemented/used.
That initial high-tech homebuilder you mentioned will only achieve the 20% reduction in the cost of construction, and won't be able to get adequate demand for the homes he gets built.
Somebody else will change the way the robotics/AI is used in the homebuilding, so that he achieves 40% reduction in the cost of construction. Somebody else will achieve 60% reduction in the cost of construction. Etc, etc.

But for these changes to happen, the robotics/AI has to at least start getting used in this sector.



The above is merely an example in automation - I only posted it because you mentioned homebuilding - but Artificial Intelligence is more about giving machines the smarts to deal with the complexities of the real world. There are related component technologies, like machine vision, optical recognition, voice recognition, LIDAR, etc, which give AI the tools it needs to grapple with the unconstrained real world.

The driverless car is an example of a machine performing a complex task that would normally require humans to perform. But of course, a driverless car can work 24/7 - it doesn't get tired, or sleepy, its reflexes don't degrade, its judgement can be continually improved just by learning from past data, and processing hardware will continue to improve each year, unlike the human brain, so that performance improves.

Stuff like machine vision may have started out being used in more controlled places like factory environments (eg. sorting parts or spotting defective ones on a conveyor belt), but obviously they're now increasingly moving out into the uncontrolled real world.

Perhaps battlefield applications will be the real stress-test that will force AI to evolve much higher levels of performance, so that it will be much more ready for ordinary everyday life. On the other hand, the more recent trends have seen consumer-level computing technologies advancing faster than military ones, because consumer applications are far more widespread and active than military ones.

But the more the technology advances, the more cases it will be open to use for, thus increasing the spread/scope of its usage, thus giving more opportunities to evolve/improve, including via machine learning.

Click to expand...

Nice story on the wonders of (future?) AI but it did not ansnwer my questions which you quoted - how does the human unable to find paying work buy anything at all?

ttwtt78640 said:

Nice story on the wonders of (future?) AI but it did not ansnwer my questions which you quoted - how does the human unable to find paying work buy anything at all?

Click to expand...

How about as robot repairmen, robot operators, robot supervisors, etc.

When computers first appeared, the predictions were that they would replace humans at all kinds of tasks, thus making lots of people unemployed, while only the big computer manufacturers profited. That didn't happen, of course.

The Gutenberg Principle happened. In the beginning, only a select few had the Bible, but one day Gutenberg invented the printing press, and soon everybody had it.

At first, there were only a few computer manufacturers, and their products couldn't really do all that much. Later on, their products got smaller and cheaper, and more people had access to them to become employed at operating them, programming them, etc. Eventually, the Personal Computer revolution was launched by none other than IBM, the original inventor of mainframe computers. Then suddenly every home and small business had a computer, and the ability to learn how to be productive on one.

Robots will go the same path. There will be more and more manufacturers, their price will come down, they'll become more available to small businesses and there may eventually be personal robots, to take care of your elderly parents or maybe your kids, perhaps with remote supervisors.

I worked at a robotics startup that wanted to implement some of these ideas. But I think they were too ambitious and too early. They wanted tele-operated robots that could be supervised remotely across the internet from a small operations center. (That's not really the same as AI, though) The number of ideas/applications were plentiful, but the industrial design & prototyping costs were prohibitive. I left there after a year, and spoke just a few days ago with a friend who's now about to leave, after being there for a few years.

There are plenty of others out there pursuing similar dreams for machines that can do a lot more than a Roomba can. It may take the deep pocketed efforts of an Amazon or an Apple to make major leaps. Apple is of course working on its own self-driving car idea, once they get it rolling and selling, you can be sure its capabilities will evolve aggressively.

iCar release date rumours | Apple Car rumours | carOS rumours - News - Macworld UK

A car would be such an excellent application for AI, because it's a very common platform which performs personally/economically important work. Human error kills so many drivers every year, and having AI in vehicles could one day save more lives than ABS brakes. And of course AI in trucks will reduce cost of transporting goods, impacting cost of living.

Companies like Nvidia, who make videogame graphics processors, will bring down the costs of AI hardware, because these graphics processors can likewise be re-purposed towards machine learning applications.

RabidAlpaca said:

The Write brothers didn't fly until 1903. We are way past the point of the most rudimentary example of the technology. We already have low level AI and it's growing exponentially. More and more jobs are being replaced as the capabilities of automation and AI grow. The time to start thinking about solutions and regulations is now, not 20 years into the problem.

We don't need to have immaculate AI to replace menial American jobs, but when we do, there will be very little need for the labor that the vast majority of humans can provide.

Click to expand...

Who are the "Write" brothers?

sanman said:

AI (Artificial Intelligence) is coming - it's already showing up in our consumer products, and it's going to continue to increasingly make itself felt in our economy. It is reducing the difficulty of doing certain jobs, and even taking over and replacing certain simple job roles that were being done by humans. There are some things machines can do better than humans - particularly repetitive tasks which require a lot of precision.

Should AI be regulated? If so, then in what ways?

As AI continues to take over and eliminate jobs from the economy, what should be done to address the unemployment that will grow from this?

Finally, should we be concerned that AI will ultimately pose a threat to humanity itself? What can we do to avoid this, while reaping the benefits of AI?

Merging AI with Human Intelligence - Should We Really Go That Far?

Click to expand...

When AI is far enough developed it will be mission accomplished for humans and they will be redundant. or is it more

sanman said:

The market tends to seek an equilibrium, whereby the desired prices of both buyer and seller intersect. So if houses become unaffordable and unbought because of the aforementioned unemployment, then their price will automatically go down. Otherwise, what's a homebuilder going to do - build a home using his robot, and just sit on it?

What will happen during all this buyer-seller price-seeking, is a change in the way the robotics technology is applied/implemented/used.
That initial high-tech homebuilder you mentioned will only achieve the 20% reduction in the cost of construction, and won't be able to get adequate demand for the homes he gets built.
Somebody else will change the way the robotics/AI is used in the homebuilding, so that he achieves 40% reduction in the cost of construction. Somebody else will achieve 60% reduction in the cost of construction. Etc, etc.

But for these changes to happen, the robotics/AI has to at least start getting used in this sector.



The above is merely an example in automation - I only posted it because you mentioned homebuilding - but Artificial Intelligence is more about giving machines the smarts to deal with the complexities of the real world. There are related component technologies, like machine vision, optical recognition, voice recognition, LIDAR, etc, which give AI the tools it needs to grapple with the unconstrained real world.

The driverless car is an example of a machine performing a complex task that would normally require humans to perform. But of course, a driverless car can work 24/7 - it doesn't get tired, or sleepy, its reflexes don't degrade, its judgement can be continually improved just by learning from past data, and processing hardware will continue to improve each year, unlike the human brain, so that performance improves.

Stuff like machine vision may have started out being used in more controlled places like factory environments (eg. sorting parts or spotting defective ones on a conveyor belt), but obviously they're now increasingly moving out into the uncontrolled real world.

Perhaps battlefield applications will be the real stress-test that will force AI to evolve much higher levels of performance, so that it will be much more ready for ordinary everyday life. On the other hand, the more recent trends have seen consumer-level computing technologies advancing faster than military ones, because consumer applications are far more widespread and active than military ones.

But the more the technology advances, the more cases it will be open to use for, thus increasing the spread/scope of its usage, thus giving more opportunities to evolve/improve, including via machine learning.

Click to expand...

Cool! no more dirty hands.

I personally believe that our inter-connectivity along with advancing AI and robotics could bring ugly consequences. At some point when robotics with advanced abilities and AI are a common theme around the world it will open a very serious security concern. Terrorist or other like minded people will not operate training camps for suicide bombers and insurgency but instead will start training camps in programming and hacking. We will move from nefarious programs that steal credit card or bank information toward programs that turn our smart technologies against us.

joG said:

Cool! no more dirty hands.

Click to expand...

I remember Dirty Hands was the title of an episode in the re-made Battlestar Galactica TV series. The chief engineer of the Galactica leaves to serve on one of the support ships, and he and his fellow workers have to live and toil in very grubby and unsafe work conditions. Ironically, that was in a sci-fi show about AI enemies (Cylons).

Robots could be great at cleaning up our environment, working 24/7 to pick up litter, etc. That was one idea I'd pitched to my boss.

Baralis said:

I personally believe that our inter-connectivity along with advancing AI and robotics could bring ugly consequences. At some point when robotics with advanced abilities and AI are a common theme around the world it will open a very serious security concern. Terrorist or other like minded people will not operate training camps for suicide bombers and insurgency but instead will start training camps in programming and hacking. We will move from nefarious programs that steal credit card or bank information toward programs that turn our smart technologies against us.

Click to expand...

Yes, I'd imagine that sooner or later someone will try to hack driverless cars, and cause some serious injury/damage. Imagine that they get very creative with their hacking, and program the car to go berserk and take as many lives as possible. Yes, like anything, AI can be misused.

I think we should define a specific part of industry for AI to take place. Also, I do not think that AI can replace humans in all fields. For example, when a translator does his job he may use some tools to simplify the translation work, but still the translation should be made by a human being who can understand the style of the translated text, the specifics of the field that relates to the text and the expectations of the customer who ordered the translation. I prefer human translation to machine translation as human translation is always more proficient than a machine one. I have business in Poland, I know Polish language a bit, and however, that is not enough for a document's translation. Machinery translation is no good in such case as it cannot translate correctly even the simplest sentences, thus, I use https://www.translateshark.com/polish.html to make sure I do not embarrass myself in front of my Polish partners. I think teachers and doctors cannot be replaced by machines as well. As when it comes to communication with people there are a lot of things that cannot be automated.

countryboy said:

Who are the "Write" brothers?

Click to expand...

That is just silly.

They are the bros that can write and spell correctly, of course. They would probably be known as the right Write Brothers. Or perhaps the Write right Brothers? :lamo

Seems to me the more advanced the society, the more quickly utilized these advancements the more competitive those societies will quickly reestablish themselves. With more efficient and lower labor costs, a USA for instance, will be in a position to compete with the lower labor cost competition already displacing, or having displaced, so many of our low skilled workers.

That, however, only solves the problem of being more competitive in the short run.

As an aside, those demanding higher and higher minimum wages are pushing this new avenue on us even faster as it becomes more and more cost effective to consider something other than a human to do easily programmable, redundant and minimal decision making tasks. This, of course, will rob those who should have those type jobs, the high school job entrant needing money and willing to take on the more menial minimum required skills jobs from learning ground up how business actually functions, acquiring new skills and starting up the ladder.

Minimum wage jobs were generally never meant for someone who is raising a family.

I do see huge future labor disruptions which in turn will cause huge market disruptions as folks do need jobs, some sort of ability to acquire some form of financing, to pay for the goods and services being developed and made in the marketplace.

Add that, with these advances there will be afforded, most probably, more and more leisure time with less and less devoted to the meaningful. This is problem needing solution.

How do we work out a world where there are good values maintained, how might we adopt a new system transferring where many of us might gain our own self worth from one of the main ones, our job, if most of us are to be replaced by something that does that job for us? If, eventually, technology provides mankind with devices that can outperform us in every single task, say for instance an art like painting, can mimic style, brush stokes of the masters, etc, beyond anything a human might be able to produce, even maybe with programmed random faults inserted so that each painting is totally unique...

What then?