Artificial Intelligence

[dragoner]

No, but while they are more efficient, they also try to not ignore the emotional needs of the people; and that is where it all dovetails together, all of it. Like "Against a Dark Background", it is where the game happens, the central aspect of the player's perception, everything is incidental to that aspect. A player also created a robot character from the Mongoose Robots book, it was interesting in that there was an exploration of servitude, and the illegality of representing one's self as a sophont. Different systems, and or polities had different ideas about it all, esp the further one got from Imperial norms.

 

[Reynard]

"I agree, I think the big question is, will these computers that are programming themselves, redesigning their hardware, firmware and software make a jump that brings them closer to a human thought process or make a quantum leap that takes them to a state that is alien to us?"

Until a thinking machine breaks the A.I. barrier, they can't improve themselves. That takes imagination, creativity, curiosity and ingenuity which is above programmed logic and processing. A computer can do things faster but they can't make leaps in creation design or much else.

 

[ShawnDriscoll]

Do you know any other way to disallow the Russians and Chinese from developing AI without our developing it first? I don't know how to do that, the best of humanity are about equal in intelligence. I don't see how you expect Luddites to rule the world before the advent of AI so they would be in position to prevent the development of AI world wide, and exercise total control to preclude hard AI. That is not a reasonable expectation, the best you can hope for is to slow us down so someone else develops the AI to its standards, and an AI developed by the Russians or the Chinese, know they are a totalitarian country, it is likelier that the AI they develop would be more likely to eliminate humanity all together, because those AI would learn from the Russians and Chinese by example rather than us. We want to influence how the AI develops, and if Luddites get their way, that won't happen.

One country developing cars or computers first does not prevent other countries from developing their own also. Serial killers are known to have high intelligence and good grades in school, etc. Not many things can be standardized. See the various languages humans and computers use. Democrats kill their own babies in the US. The Chinese do as well in their country. An AI could mimic such standard of operations.

We can simulate intelligence to possibly a superior one to a human intelligence, because a computer can process information and possible outcomes a lot faster than we could.

We can once we figure out how brains work, which no one has been able to yet.

Who's to say we can't, it is a complex task to analyze and simulate the brain, but there is nothing to suggest that it is impossible.

Just you and me typing here is causing errors. "Once we figure out how brains work" does not mean "it is impossible". To even make an AI, people on the same team have to be able to communicate with each other first.

At the very first instance of some AI doing something out of line, like breaking laws, I think people would "discipline" the computer. Try to teach it that it is wrong. If necessary "ground" it. Take away their cell phone and internet access until they "promise to behave". Maybe send it to "rehab". Perhaps remove or rewrite code and data the AI created. A rebellious enough AI, that repeatedly breaks laws would be sent to jail for life. Shut down.

Depends on the culture. Some countries would rather medicate than arrest/execute criminal types.

 

[ShawnDriscoll]

I don’t see the problem, money would work in a similar way that it does now but the value chain would be very different. Different economic activities could be run by different AI subsystems that would trade with each other just as businesses trade with each other today. At least, that’s one possible model.

Too vague. The stock market is just a betting game, not connected with anything outside of it other than with paying out to "investors".

The difference is that today there is a limit on the total amount of work that humanity can do each year, and a low limit on how fast that work capacity can grow. The growth rate is determined by population growth combined with productivity multipliers such as technological development and infrastructure improvement.

A huge factor is government, which puts limits and regulations on every aspect of business and the economy engine.

An AI driven world changes the labour force growth rate from a steady few % per year to a drastically higher rate.

Too vague. What does "AI driven" mean exactly? Where will the money come from to get the factories built, to get the materials needed to make the AIs, to manage the AIs who will working in those factories...? And so on.

In a single human generation instead of the labour force growing by maybe double at most, with AIs manufacturing ever improving worker robots it could increase a hundred fold in the same time span. Humans would not be able to contribute significant labour, and capital accumulation would no longer be a problem because the cost of capital equipment would be super-low.

You cannot take gold and pound it or stretch it to make more gold.

In a world in which the cost of labour is 1% of what it is now and the capital investment needed for a project is also 1%, the only real question is how would we choose to use the resources available to us. Resource utilisation and environmental impact become the constraining factors on economic activity, so that’s where the value would be.

Where does the investment come from? How much money does the government take from the profits made? What business would have that kind of money as capital to start with? Would the business ever see a return from all that effort? Assuming prices are kept low for humans (who are now jobless) to buy things, will any profits be enough to pay for all these AIs being built/maintained/upgraded/refurbed/DRMOd? Remember, the government is taking a huge amount of the company's money (a perverse Robin Hood mindset still in effect there).

Ok. Either AIs are controlled/regulated by humans, or they are not. Which is it?

I’m assuming they will be controlled and regulated by humans that set targets and define the parameters within which the AIs operate. I’m not aware I’d suggested otherwise. That’s not guaranteed of course, but I think the alternative is an essentially unknowable future that would be hard to speculate about meaningfully.

Well, anything controlled by a committee goes corrupt. And there is no accountability. No one gets fired or sent to jail. So this AI project is just a scheme to tax people more. If not a tax, it is simply called a fee instead. A mandated one. All humans are charged a fee just for being alive, to help pay for the AI project. The money ends up going into politician pockets and the AI takes years to think of something on its own.

 

[Tom Kalbfus]

Where does the investment come from? How much money does the government take from the profits made? What business would have that kind of money as capital to start with? Would the business ever see a return from all that effort? Assuming prices are kept low for humans (who are now jobless) to buy things, will any profits be enough to pay for all these AIs being built/maintained/upgraded/refurbed/DRMOd? Remember, the government is taking a huge amount of the company's money (a perverse Robin Hood mindset still in effect there).

Where it usually comes from in all such ventures, investors. How much of a percent would need to be taken? Ais have the capability to reduce costs alot, that small amount that gets redistributed, after deflation can be made to be enough. Competition between low cost producers will tend to reduce prices, so a pittance can buy a lot, we just need the government to redistribute that small amount to people who can't find jobs and were done!

Problem is if their A.I. is designated Guardian and we name ours Colossus and they decide to chat and they call Skynet for a second opinion...

A.I.s are thinking machines. Like people, they don't always do what mom and dad taught them. Rebellious cyberteens!

Well that means we have to get those AIs right now doesn't it. Fortunately we have some experience raising children, and most of those children turn out all right, though a few are psychopaths, We just have to make sure we raise our AI "children" right, and if we get a good one, we can copy it as many times as we need! Also parents have a lot to do with how a child develops, if a parent is abusive then the child is more likely to pick up the parents trait.

If you watched the movie Ex Machina, you'd see that the "daughter" (in the center) picked up personality traits of her "father" (on the right). The Robot's creator was a ruthless Frankenstein type, he liked to play mind games, and of course his robots picked up from him and ended up killing him to escape confinement. If we are to develop AIs we have to make sure they have the right "parents" the example in this movie shows what can happen if they have the wrong "parent".

 

[ShawnDriscoll]

And rightly so, you make the point it seems many are missing here: that's the point of sci-fi and sci-fi RPGs to speculate about this future. Who said anyone had to get it right?

That is the number one issue with tabletop RPGers. They want to make the leap, the connection, from our present setting to the game's setting. For the game to have its society and technology, so that of course we in the present (maybe a few years from now) are doing something so that the timeline listed on page 18 of the rules in the game matches perfect with what's happening soon in our own time. This is part of the beauty of when marketing an RPG to a 14-34-year-old male demographic. Because most players just play themselves in these games. So they need to make the leap from here to the game's there.

The challenge for me is how do you "play" (I'd keep them as NPCs) a computer that's evolved? Personally I don't see them getting hung up in emotional sh!t as all humans do, that people confuse emotional responses with intelligence is just depressing.

The trick there is to use traits for your AI characters. I think the Traveller charsheet has a Traits or Racial Traits field on it. Just fill that in with the differences that only an AI would have. Then role-play, using those traits from time to time, to remind the human player characters that someone in the room is not also human, as they would think.

Now what I do like is expert systems that run things, the internet of things as we are seeing develop today. That way my space ship will always order fuel and life support the moment it drops out of jump and before the "crew" have to decide what to have for breakfast, the ship has gone thru all the protocols with SPATC and a berth is being prepped by robots controlled by another super fast super diligent and overal benevolent computer.

That way my players can get on with messing things up with other humans like they always do...

I've always used computers with cool UIs and have expert systems running on them in my games. My AIs tend to work the same as those found in Gene Wolfe books.

 

[ShawnDriscoll]

Where does the investment come from? How much money does the government take from the profits made? What business would have that kind of money as capital to start with? Would the business ever see a return from all that effort? Assuming prices are kept low for humans (who are now jobless) to buy things, will any profits be enough to pay for all these AIs being built/maintained/upgraded/refurbed/DRMOd? Remember, the government is taking a huge amount of the company's money (a perverse Robin Hood mindset still in effect there).

Where it usually comes from in all such ventures, investors.

Those days are long gone now. Those kinds of jobs/businesses are long gone now, other than the car and plane manufacturers that were already around before the big boot on our necks. The only way to start a business now is to get money from the government (or from a bank which is controlled by the government). And the government just prints its own money, so it doesn't have to compete like a business has to.

How much of a percent would need to be taken? Ais have the capability to reduce costs alot, that small amount that gets redistributed, after deflation can be made to be enough. Competition between low cost producers will tend to reduce prices, so a pittance can buy a lot, we just need the government to redistribute that small amount to people who can't find jobs and were done!

So basically, the government is running the AI business. So this AI is not going to happen. Not when there are career government employees in the loop, slushing the money around to various departments and hiding the numbers from auditors.

 

[Condottiere]

As I recall, there is a programme to pick out likely pop hits, and one to write them, though the interviewee was very cagey as to who his clients were; I suspect Swedes.

 

[Tom Kalbfus]

Where does the investment come from? How much money does the government take from the profits made? What business would have that kind of money as capital to start with? Would the business ever see a return from all that effort? Assuming prices are kept low for humans (who are now jobless) to buy things, will any profits be enough to pay for all these AIs being built/maintained/upgraded/refurbed/DRMOd? Remember, the government is taking a huge amount of the company's money (a perverse Robin Hood mindset still in effect there).

Where it usually comes from in all such ventures, investors.

Those days are long gone now. Those kinds of jobs/businesses are long gone now, other than the car and plane manufacturers that were already around before the big boot on our necks. The only way to start a business now is to get money from the government (or from a bank which is controlled by the government). And the government just prints its own money, so it doesn't have to compete like a business has to.

How much of a percent would need to be taken? Ais have the capability to reduce costs alot, that small amount that gets redistributed, after deflation can be made to be enough. Competition between low cost producers will tend to reduce prices, so a pittance can buy a lot, we just need the government to redistribute that small amount to people who can't find jobs and were done!

So basically, the government is running the AI business. So this AI is not going to happen. Not when there are career government employees in the loop, slushing the money around to various departments and hiding the numbers from auditors.

What do you think Google is doing with its self-driving cars? IBM is funding project Blue Brain, the money comes from both private and public sources. Just watch what happens in the next two years, I think it reached a tipping point.

http://pics.mcclatchyinteractive.com/wire_photos/4n4ptg/picture27641566/ALTERNATES/FREE_960/GOP%20Trump%202016.JPEG[/img]
Just watch these people over the next two years. There is a phenomenon going on here.

 

[ShawnDriscoll]

What do you think Google is doing with its self-driving cars?

I think Google has a lot of money of its own to try new things with. Will it be allowed to expand? Or will the government make it a money loser for it or for anyone trying to copy the idea after Google runs out of money developing the technology? Businesses succeed when they are producing something that people actually need. The cost to keep each car running is probably high. Lots of engineers will be involved behind the scenes on each one. More-so than with regular taxi company cars. I'm sure nobles and dilettantes, who can afford their service, will be saying on their red carpets that everyone should be using them to save the ..., etc.

I looked up Iain Banks. He lived in Scotland. He was for its independence from the UK, so his politics were far left. And so he had to hand-wave all the utopias used in his sci-fi stories in order for them to work (see also Gene Roddenberry's Star Trek and other TV shows he did, as well as George Lucas' stuff). Which is fine. It's just that a lot of his readers think that his utopias are sound and are possible, and/or it will be our future, etc. Unfortunately, his stuff is just sci-fi-fantasy.

 

[simonh]

I don’t see the problem, money would work in a similar way that it does now but the value chain would be very different. Different economic activities could be run by different AI subsystems that would trade with each other just as businesses trade with each other today. At least, that’s one possible model.

Too vague. The stock market is just a betting game, not connected with anything outside of it other than with paying out to "investors".

I'm losing patience here. I didn't mention stock prices or stock trading either directly, indirectly or even by inference. I'm talking about buying and selling commodities, products and services between businesses or economic entities. Corporate ownership structures have literally nothing to do with that. Even socialist systems like pre-Deng Xiaoping China and the Soviet Union contained business entities that traded with each other using money, but had no concept of stock ownership or stock trading. I have no idea what your statement here has to do with what I wrote. Looking at some of your other responses, I see a similar fundamental disconnect between what I'm writing about and your commentary on it, to the stage where it's hard to see any point in responding.

Simon Hibbs

 

[ShawnDriscoll]

I'm losing patience here. I didn't mention stock prices or stock trading either directly, indirectly or even by inference. I'm talking about buying and selling commodities, products and services between businesses or economic entities.

If you're talking about companies outsourcing, or contracting with other companies, or becoming tech-partners, or doing trading with each other, or how companies run their procurement departments (again, I can't read your mind), I don't see how an AI would improve anything except for maybe cutting out the human beings involved. But for every AI running a department, there are 5-6 IT guys involved. So it ends up costing more to run those departments. Governments look to business to do these things for them because governments are so top-heavy and inefficient. Their idea is to keep throwing more AIs at a project until it is completed.

 

[simonh]

What do you think Google is doing with its self-driving cars?

I think Google has a lot of money of its own to try new things with. Will it be allowed to expand? Or will the government make it a money loser for it or for anyone trying to copy the idea after Google runs out of money developing the technology? Businesses succeed when they are producing something that people actually need. The cost to keep each car running is probably high. Lots of engineers will be involved behind the scenes on each one. More-so than with regular taxi company cars. I'm sure nobles and dilettantes, who can afford their service, will be saying on their red carpets that everyone should be using them to save the ..., etc..

Commercial self-driving vehicles are already licensed and in operation on the roads. It's still a trial but these are not prototypes, they are fully propduction ready systems already licensed for road use.

http://arstechnica.co.uk/cars/2015/05/freightliner-unveils-the-first-road-legal-self-driving-truck/

Note thet these are commercialy viable even though the vehicles currently still carry a driver that supervises the system. Autonomous driving systems are cheaper to run on an ongoing basis because their driving behaviour puts less stress on the vehicle's mechanical systems, so they last longer, and are considerably more fuel efficient than human drivers. Insurance costs will eventually also be lower because the accident rate is lower (based on millions of miles of autonomous driving data from test systems).

Simon Hibbs

 

[ShawnDriscoll]

Note thet these are commercialy viable even though the vehicles currently still carry a driver that supervises the system. Autonomous driving systems are cheaper to run on an ongoing basis because their driving behaviour puts less stress on the vehicle's mechanical systems, so they last longer, and are considerably more fuel efficient than human drivers. Insurance costs will eventually also be lower because the accident rate is lower (based on millions of miles of autonomous driving data from test systems).

Cheaper to run than what?
They last longer than what?
More fuel efficient than human drivers?
Accident rate is lower than what?

All platitudes. Uber is sweating right now how it will survive, now that the government has breathed over them. Just a matter of time before Google decides which countries are best to make money in before having to sell their technology to someone as a write-off. Lucas had to sell off his Star Wars to save on California taxes being aimed at him.

Anyway, test-track data does not relate at all to real-world use. Fantasy tech counts chickens before they're hatched, etc.

 

[simonh]

I'm losing patience here. I didn't mention stock prices or stock trading either directly, indirectly or even by inference. I'm talking about buying and selling commodities, products and services between businesses or economic entities.

If you're talking about companies outsourcing, or contracting with other companies, or becoming tech-partners, or doing trading with each other, or how companies run their procurement departments (again, I can't read your mind),

No, what I said was that AIs would run all economic activities. Design, resource extraction, manufacturing, distribution, retail, everything. Not just things like outsourcing.

I don't see how an AI would improve anything except for maybe cutting out the human beings involved. But for every AI running a department, there are 5-6 IT guys involved. So it ends up costing more to run those departments. Governments look to business to do these things for them because governments are so top-heavy and inefficient. Their idea is to keep throwing more AIs at a project until it is completed.

AIs wouldn't need IT guys because they would be capable of doing their own IT. You're carrying around a lot of assumptions and pre-conceptions that just don't apply in a world with pervasive AI.

Simon Hibbs

 

[ShawnDriscoll]

AIs wouldn't need IT guys because they would be capable of doing their own IT. You're carrying around a lot of assumptions and pre-conceptions that just don't apply in a world with pervasive AI.

Clarity is everything. I see where you are coming from now. You're assuming that there will be AIs running everything. Hand-wavium technology.

 

[simonh]

quote]
Cheaper to run than what?
They last longer than what?
More fuel efficient than human drivers?
Accident rate is lower than what?

Vehicles driven by human drivers.

All platitudes. Uber is sweating right now how it will survive, now that the government has breathed over them. Just a matter of time before Google decides which countries are best to make money in before having to sell their technology to someone as a write-off. Lucas had to sell off his Star Wars to save on California taxes being aimed at him.

Anyway, test-track data does not relate at all to real-world use.

Google's self-driving cars have logged 2 million miles of operation on real public roads, not test tracks. They use them for street view, which is a commercial purpose so in fact self driving cars have been used actively for commercial purposes for several years now. They are only one self-driving car developer.

Uber's problems are all down to the fact that they have human drivers. Once they switch to driverless cars all the employment regulation issues dissapear.

Simon Hibbs

 

[simonh]

AIs wouldn't need IT guys because they would be capable of doing their own IT. You're carrying around a lot of assumptions and pre-conceptions that just don't apply in a world with pervasive AI.

Clarity is everything. I see where you are coming from now. You're assuming that there will be AIs running everything. Hand-wavium technology.

For now it's hand-wavium. As I've said before, we don't have any idea how to actualy build a true 'strong' AI right now. However if we do eventually do it, by definition thy'll be able to do everything that we do at least as well as we do it, and probably an awful lot more an awful lot better. Once we have superior AIs, there will be no work that humans can do that the AI's won't be able to do better and more efficiently (and therefore, more cheaply). That won't be true of the first geenration of AIs, they'll live in huge expensive data centres and concentrate on the highest value activities, but as the technology matures AIs will take over more and more economic activity.

I don't think they will necessarily _literally_ take over everything completely, there will be edge cases. There are many economic activities we perform because we enjoy them.

Simon Hibbs

 

[Reynard]

"Autonomous driving systems are cheaper to run on an ongoing basis because their driving behavior puts less stress on the vehicle's mechanical systems, so they last longer, and are considerably more fuel efficient than human drivers. Insurance costs will eventually also be lower because the accident rate is lower (based on millions of miles of autonomous driving data from test systems)."

Sounds like we need to take a step back. Remove all crews from ships and trains and completely automate them. Today's technology is more than capable plus rail and shipping lines should be a lot less complex to navigate. Corporate heads won't know what to do with all that money!

 

[simonh]

Sounds like we need to take a step back. Remove all crews from ships and trains and completely automate them. Today's technology is more than capable plus rail and shipping lines should be a lot less complex to navigate. Corporate heads won't know what to do with all that money!

It will also make a lot of things much cheaper for us consumers too once operating with no driver supervision gets approved. The down side is that it will wipe out whole swathes of reasonably well paid jobs. Not just drivers themselves their families and the communities they live in, but also a vast infrastructure of truck stop cafes, restaurants, motels and local stores that depend on revenue from truck drivers passing through. Huge tracts of the midwest are heavily dependent on this side of the transport sector. I'm actualy a Brit and the collaterial social impact of this kind will be less severe over here but it will have an effect, which will play out over the next couple of decades.

I'm not involved in the transport sector at all, so I have no skin in this game and only stand to benefit from falling costs, but I can't ignore the fact that the negative impact will be huge on a lot of people. When a new technology comes alonmg everyone says it will have huge impact in 3 to 5 years. It's a terrible cliche. But in this case I think it's actualy true. These things exist on our roads right now. At this point it's just a matter of regulation and deployment capacity.

Simon Hibbs