"Software (Bot) Substituion" A Gorean Forums Think-tank

User avatar
Theoden
Posts: 398
Joined: Wed Oct 05, 2011 6:20 am
SL Name: Theoden
Caste: Warrior

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Theoden » Mon Mar 17, 2014 5:06 pm

Those are predictions 40 to 50 years into the future from that point in time.

I'm not saying AIworld won't happen, I just don't think it's going to happen in the next 10 to 20 years.
User avatar
Anarch Allegiere
Posts: 1006
Joined: Thu Aug 26, 2010 8:01 am
SL Name: Anarch Allegiere

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Anarch Allegiere » Mon Mar 17, 2014 6:35 pm

Predictions are pointless, sometimes they hit, sometimes they miss.

Reminds me of the financial experts of the US government estimating that Facebook was never going to be worth more than a few thousand dollars.

There seems to be quite a bit of technological slowdown going on though. There's several predictions, or rather well-informed estimates, I remember from around the year 2000. The estimate of many in the research field that Fusion Reactors (Tokamak etc.) would be providing residential electricity by the year 2010, yet still nowhere in sight in the year 2014.

Or the more related one... that an AI robot soccer team would be able to defeat the best human team in a game consistently by the year 2012. This one disappoints me the most in not yet having achieved that goal...
User avatar
Glaucon
Posts: 2832
Joined: Tue Nov 03, 2009 9:07 am

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Glaucon » Mon Mar 17, 2014 7:59 pm

Hawt Sommer wrote:Asimo AI learning...
http://www.youtube.com/watch?v=QfPkHU_36Cs


The last part of that item is indeed the really interesting thing. It also indicates that real (turing test proof) AI is still a long way off, as well as indicating that steps on the road towards it are being taken. Baby steps, perhaps, though some people boldly claim that they are working on projects that will make big leaps.
User avatar
Glaucon
Posts: 2832
Joined: Tue Nov 03, 2009 9:07 am

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Glaucon » Mon Mar 17, 2014 9:40 pm

Some point by point replies to Théoden (and Anarch, later on).

Theoden wrote:Glaucon, I think what you say will be relevant in 50 to 60 years time in Western countries, but you are being very bold to suggest it might happen in 10 years.


I think I was saying 20 years. That is the number usually tossed out by economists and reporters talking about this potential issue, lately.

It's 2014. Think back to 2004. Did things change all that much? Some things yes. Smart phones, some digitization of business practices and finance, internet usage. But not catastrophically, as you paint it, an entire digital revolution.


Think back to 1994. Did things change by a lot, since then? Well, yes. The internet happened, pretty much (it did already exist before that, but it became a big thing right about then).

What happened in offices is something I'd say happened between roughly 1985 till 2005: All or nearly all primary work-related information got to be stored and accessed into computer programs. Of course, paper files were usually still kept. But these paper files (in so far as they still used) are now usually a 'double' for stuff already in a computer system.

The problem with most of that information in those computer programs that, very often, it exists in separate systems that cannot really talk to each other. So, you still need humans to make these systems (within organizations and between organizations) talk to each other: either through simple means (people printing out information and entering it manually into a new system) or in a 'smart way' (people using software - like excel or other programs) to import and export mass data in and from different systems, or more expert people ensuring that systems can talk to each other directly. But now that nearly every commercial organization is fully linked to the internet, with more and more companies being tempted to store their data 'in the cloud', it is possible for people to start adopting universally usable data for many things.

For example: Nearly every company makes invoices for stuff they sell, and receives invoices for stuff they buy. Typically, they send out invoices on paper and receive them on paper. Tax authorities often want companies to have these things in a paper format, still. But what if this sort of information was send and received through a universally accepted code (possibly supplied by a third party, like a conglomerate of banks)? Then an invoice could just be automatically read into a system, automatically send through an internal approval process. Payment of that invoice could be confirmed automatically. Invoice payment chasing could be done automatically at first. So, instead of the accounting dep. of that middle-sized company having a few people making invoices and sending them out, and entering received invoices and readying payment for them, you'd just need one guy to keep an eye on it, to deal with the odd things and to worry about the 'higher end' sort of accounting stuff (like keeping an eye on individual and overall spending, cash management, and so forth). The 'lower-level' office work would be drastically reduced for that department, with just the higher level work remaining.

The same would be true for many 'office departments'. Because, let's face it, a large part of what the people working in big offices do is paper-pushing. I know what you might be thinking: wasn't this supposed to have happened before? Like the 'paperless office'? Yes. Things take time. But Ten years ago, a lot of companies weren't fully connected to the internet yet. Now they are, and now these kind of changes are quite possible. Maybe it will still take some time before really innovative software starts connecting the dots (most of the stuff a lot of companies are working with looks like it is still made with MS-DOS with a bit of windows window-dressing on top of it), but that can't last long, I think. At some point, there will be a 'facebook' of business administration.

People are slow to adapt. These technologies require a lot of money to invest in.


Do they? In the past (still, perhaps), companies tended to try and have their 'own' system. Custom made programming, modular stuff customized for that company, and so forth. That is usually why many companies have a zillion different applications running that aren't designed to work with one another. But companies also use standard software. Like Microsoft Office. Whatever you think of it, that stuff does work. Often, they communicate THROUGH office program files (word documents, excel files, outlook mails and so forth). Sharing isn't just a thing of consumers, it is important to businesses as well. I expect that companies will increasingly stop trying to have their own individual customized software for this and that, and will increasingly start using 'ready made' solutions, because, over-time, customized systems tend to lead to inefficiency.

Time for people to learn these technologies as well. Many companies and institutions just don't have the capital to plunge into these things. Maybe some, but not all.


Ready-made solutions sold to many companies and people would be relatively cheap (again, like Office) and they would have another important advantage. People don't stay with companies for as long as they used to, in the past. It is very costly to have to teach new people how to use custom-made or company specific software. With more job mobility, people would know standard software, usually. Also, if you think in terms of 20 years, that is nearly a generational shift. Sure, people do resist change, but new people aren't used to the old ways of doing things (though they can get used to it, of course). And management can be slow to adapt. But the management of a company shying away from change for twenty years? That usually spells a bad future for such a company. And in twenty years, that is usually enough to make that company disappear (bankruptcy or, more likely, that company being sold to a bigger one that does better).

A lot of the time, when new technology comes out, people just sit and wait until other people have tried it and proven it's profit maximization and cost/benefit projection. This takes time. The entire economy won't become this AI world in the next decade.


Like I said... 20 years was what I had in mind. That is not a short time. 20 years ago, there were still ashtrays on the desks in offices. And a lot of people were still getting their first monitor (perhaps still one of those with green text only).

If you think competition will force people to invest in new technologies, then you underestimate the barriers to entry of companies into markets.


Right. Twenty years ago, no one had heard of Google, and Apple was struggling. IBM was still the IT giant. Sony was still the crème de la crème of consumer technology. And what could ever happen to General Motors and Chrysler?

There are a lot of companies that have a monopoly or have few competitors, just because other competitor's aren't big enough or have enough reach in their area.


Sure. British Petrol will probably still be British Petrol, even if they aren't very innovative. But companies like that DO have internal competition. If one branch in one country can reduce costs by using particular methods, the overall management will be interested and probably want other branches to do the same thing. This won't always work. But 20 years is a long time, including changes in management, usually.

Even now, smaller businesses and medium businesses exist in areas simply because the customer is too lazy to economically maximize utility-cost, even though they are selling inferior products at higher prices. People will go to small and medium businesses just because it's close by. Maybe some big companies just don't deliver into that area. And I highly doubt Amazon's 'Delivery by Drone' thing is going to work... it's too easy to shoot things down and steal shit. Maybe even jamming electrical control systems of the drones for 'funsies' as the Iranians did with our CIA drone a few years back. It's not that difficult, and installing countermeasures is costly. That's an example of technology vs working men. UPS will win out, Amazon with their drones won't. At least until they come out drone trucks with ground drone delivery robots.


I don't know about Amazon's 'drones'. But I do know that many local bookshops are struggling, many going under. I usually by my books online, these days, too. Because the selection is much larger, and because the books are cheaper. Record shops? Hah! Sure, people still like going shopping. But competition from online shops is really hurting a lot of retail business. Even for stuff that people generally seem to want to check out in real life. People go to the shoe store and look at what they have, maybe try it on. And then they go online and order it there, because it is 20 dollars cheaper, or because they wanted it in a color that the shoe shop didn't have.

My point: Competition as a driving force into investing into the latest technologies as soon as they come up, is overemphasized. Some big corporations, yes, some startups (most startups fail) but not all will plunge into AI technology. Not the entire economy.


The 'entire economy' is rather big. But I think you are wrong. In theory, companies can still operate without using a single computer (for the most part, at least). And yet, nearly every company that is bigger than just one or two people these days does use computers. About 30 years after they became available and affordable for large-scale commercial use. I think that says something about the speed of innovation. Like I said, 20 years is not a short time. I do not buy into the often-told story about how innovation is constantly 'speeding up', necessarily. But I don't think implementation of innovative methods and technologies is slowing down either.

Safety is another thing. Anything digital that is connected to the internet can be hacked. It just takes time and resources. But there is not a single encryption that can't be broken with enough time and resources. If you were to connect all public transport to a computer mainframe with automatic systems, with a click of a button you can just have all the buses go and drive off a cliff?


And has that ever stopped anyone? Credit cards are really unsafe (for the banks, anyway) but were and are widely used. Any code can be broken, and yet nearly all of us (as well as companies) are happy to handle something very important (their money through their bank accounts) online, through systems we know aren't completely safe.

And can we really trust an automated bus driver? Well, we KNOW that human bus drivers sometimes drive off cliffs. What if the statistics show that the automated bus drivers get into far fewer accidents? Airplane pilots usually let the autopilot handle most of the flying, and that seems to lead to fewer accidents. An automated system won't get distracted by something, will have a very short reaction-time and (I'll assume) won't drive drunk or fall asleep behind the wheel. I think that this 'fear' of handing over control to a machine is something of a myth when it comes to practical stuff. People seem perfectly alright handling control over to machines when it is convenient.

At least, if you want to do that in todays world, you would have to infiltrate real humans into becoming suicidal bus drivers. But in this new digital world you paint, any shadowy terrorist organization with enough computer resources anywhere in the world can cause mass havoc and mayhem. Digital defense requires a constantly vigilant staff of security experts to monitor the situation. This is expensive. So what happens when the entire economy is linked to technology, an 'Internet of Things', and there's only a few security experts coalesced in a few areas of society, like defense, power plants, a few major city's mass transit systems.... what about the rest? What about smaller cities and their infrastructure?


Eh? Most of what you mention is the case, already. Which is why all these gov. security bigwigs wanting more budget are always going on about cyber terrorism.

When you recite this digital world, all I can think of is dystopia.


Hmmm, yes, that imagine does come up (see the other thread).

A vastly in-equal society as such we've never seen, of financial and IT overlords becoming the 1%, and 50% of the country becoming the new poverty. Superrich and Proletariat. Except they wouldn't really be proletariat, because for them to be proletariat, they would have to be working men, wouldn't they? But no, the robots are there in the factories.


Robots and automated systems becoming the police, control by drones by someone behind a computer screen somewhere... Have you seen the movie Elysium? That is what I picture, when you paint the above mentioned picture. Massive societal inequalities, with the ruling class with the keys controlling technologies to oppress the masses.


Seen the beginning, yes. Not my prediction of what the future will be like.

So much power, able to be controlled by so few. The loss of privacy. The constant surveillance. It would only take one catastrophe to happen, with a group rising, promising security at the cost of freedom, and then we would all be living under tyranny and be enslaved by a quick digital takeover.


Hmmm... well... not so sure about the 'take-over'. ;)

And then, only the second amendment would save us all. And that's assuming judges haven't gotten rid of it yet by creative 'interpreting' of the constitution.


Right. *Cough* Rand Paul for President! :mrgreen:

Can't say I really agree with your analysis. One way or another, in 10 years, in 20 years, in 50 years, a hundred, part of this stuff will come true. Resisting it with a gun left over from the civil war hidden under your grandmother's bed won't stop that. You Americans (yes, I am generalizing) seem to have this romantic fantasy of the future of a dystopia in which you all fall back on pioneer-like individualist 1776-resistance-mode and in which the small community 'fights the power' (and wins). But realistically, in such a struggle, Mat Damon and Neo and all those other resisting types would actually lose, however heroic their fight. If this image of the future has any validity, it will be because people will embrace it willingly. Everyone will want to take the blue pill, instead of the red, to use Hawt's terminology. And those insisting on the red will be labeled crazies and terrorists. And who is to say that the truth of those taking the red pill is the real truth?

Anarch wrote:Predictions are pointless, sometimes they hit, sometimes they miss.

Reminds me of the financial experts of the US government estimating that Facebook was never going to be worth more than a few thousand dollars.


But at least one person did figure that Facebook was going to pretty big: Mark Zuckerberg. I guess that, for him, that prediction wasn't entirely useless. ;)

There seems to be quite a bit of technological slowdown going on though. There's several predictions, or rather well-informed estimates, I remember from around the year 2000. The estimate of many in the research field that Fusion Reactors (Tokamak etc.) would be providing residential electricity by the year 2010, yet still nowhere in sight in the year 2014.

Or the more related one... that an AI robot soccer team would be able to defeat the best human team in a game consistently by the year 2012. This one disappoints me the most in not yet having achieved that goal...


It all depends on what you expect. When you look back at predictions about progress from the past, you see that many do not come true. Some do. Some things come true that were predicted by few. So you should not just look at those past predictions in judging progress in general.

But I do agree that, in some areas, progress is slower than I would have expected, when I was a younger man. Some things turn out to be quite difficult. Like tennis-playing artificially intelligent robots and nuclear fussion. And I certainly do not buy into the claims of the prophets of progress that say that everything is going to change more quickly all the time.

Regardless, whether it goes more quickly or more slowly, technology is constantly evolving and changing our world. I don't see that changing any time soon.
User avatar
Theoden
Posts: 398
Joined: Wed Oct 05, 2011 6:20 am
SL Name: Theoden
Caste: Warrior

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Theoden » Mon Mar 17, 2014 11:19 pm

Glaucon wrote:
Theoden wrote:Glaucon, I think what you say will be relevant in 50 to 60 years time in Western countries, but you are being very bold to suggest it might happen in 10 years.


I think I was saying 20 years. That is the number usually tossed out by economists and reporters talking about this potential issue, lately.


Ah, sorry I must have missed that.



Glaucon wrote:
Prime Bbcode Spoiler Show Prime Bbcode Spoiler:
Think back to 1994. Did things change by a lot, since then? Well, yes. The internet happened, pretty much (it did already exist before that, but it became a big thing right about then).

What happened in offices is something I'd say happened between roughly 1985 till 2005: All or nearly all primary work-related information got to be stored and accessed into computer programs. Of course, paper files were usually still kept. But these paper files (in so far as they still used) are now usually a 'double' for stuff already in a computer system.

The problem with most of that information in those computer programs that, very often, it exists in separate systems that cannot really talk to each other. So, you still need humans to make these systems (within organizations and between organizations) talk to each other: either through simple means (people printing out information and entering it manually into a new system) or in a 'smart way' (people using software - like excel or other programs) to import and export mass data in and from different systems, or more expert people ensuring that systems can talk to each other directly. But now that nearly every commercial organization is fully linked to the internet, with more and more companies being tempted to store their data 'in the cloud', it is possible for people to start adopting universally usable data for many things.

For example: Nearly every company makes invoices for stuff they sell, and receives invoices for stuff they buy. Typically, they send out invoices on paper and receive them on paper. Tax authorities often want companies to have these things in a paper format, still. But what if this sort of information was send and received through a universally accepted code (possibly supplied by a third party, like a conglomerate of banks)? Then an invoice could just be automatically read into a system, automatically send through an internal approval process. Payment of that invoice could be confirmed automatically. Invoice payment chasing could be done automatically at first. So, instead of the accounting dep. of that middle-sized company having a few people making invoices and sending them out, and entering received invoices and readying payment for them, you'd just need one guy to keep an eye on it, to deal with the odd things and to worry about the 'higher end' sort of accounting stuff (like keeping an eye on individual and overall spending, cash management, and so forth). The 'lower-level' office work would be drastically reduced for that department, with just the higher level work remaining.

The same would be true for many 'office departments'. Because, let's face it, a large part of what the people working in big offices do is paper-pushing. I know what you might be thinking: wasn't this supposed to have happened before? Like the 'paperless office'? Yes. Things take time. But Ten years ago, a lot of companies weren't fully connected to the internet yet. Now they are, and now these kind of changes are quite possible. Maybe it will still take some time before really innovative software starts connecting the dots (most of the stuff a lot of companies are working with looks like it is still made with MS-DOS with a bit of windows window-dressing on top of it), but that can't last long, I think. At some point, there will be a 'facebook' of business administration.


It's one thing to say in 20 years between 1994 to 2014 we have advanced office data systems by using computers and the internet, and in the next 20 years we'll see Big Data come around and a standardization of mass marketed business software.

It's another to say in 20 years our entire economy will be transformed with highly mechanized AI robots, able to be economical enough and advanced enough mechanically and software wise to do a variety of manual labor type things such as flipping the burgers, cleaning the floors and bathrooms, gardening, farming and the like. These may be simple tasks for a human, but for a piece of machinery, these things are actually quite complex. Walking, for instance, is insanely complex to create artificially.

Understanding humans and human interaction, and conducting creative problem solving solutions for anomalous problems, will probably be the biggest hurdle, and making it all economical that it's worth buying these robots to do these things will be the second. Maybe in 50 years, but not in 20 for the vast majority of the economy. Maybe in a city or two, as a novelty. Not across the entire country. I'd imagine it to be something like the movie I, Robot with Will Smith.




Glaucon wrote:
Time for people to learn these technologies as well. Many companies and institutions just don't have the capital to plunge into these things. Maybe some, but not all.


Ready-made solutions sold to many companies and people would be relatively cheap (again, like Office) and they would have another important advantage. People don't stay with companies for as long as they used to, in the past. It is very costly to have to teach new people how to use custom-made or company specific software. With more job mobility, people would know standard software, usually. Also, if you think in terms of 20 years, that is nearly a generational shift. Sure, people do resist change, but new people aren't used to the old ways of doing things (though they can get used to it, of course). And management can be slow to adapt. But the management of a company shying away from change for twenty years? That usually spells a bad future for such a company. And in twenty years, that is usually enough to make that company disappear (bankruptcy or, more likely, that company being sold to a bigger one that does better).



It's one thing embracing digital versions of information databases. It's another to expect artificial intelligence to entirely change the experience of the consumer as well as how you deal with every day life.

Right now I can only imagine automated trains and automated thermostats being in the next 10 years, and automated cars on a nationwide scale in the next 20 years. Definitely not on the scale as you present, with the entire lower-tier and lower middle tier job market being wiped out by robots and automation.

Glaucon wrote:
A lot of the time, when new technology comes out, people just sit and wait until other people have tried it and proven it's profit maximization and cost/benefit projection. This takes time. The entire economy won't become this AI world in the next decade.


Like I said... 20 years was what I had in mind. That is not a short time. 20 years ago, there were still ashtrays on the desks in offices. And a lot of people were still getting their first monitor (perhaps still one of those with green text only).



What does smoking cultures and regulations and a better version of the same technological concept (a monitor without green text only to a flat monitor with pretty colors and graphics) have to do with low adaptation of radical ground up technology that they would sit and wait until other companies have proven its profit maximization and cost/benefit worthiness?

Better editions of same technologies (better computers) is minor.

Bringing A.I. and drones in such a large scale that wipes out tens of millions of jobs is another.

Glaucon wrote:
If you think competition will force people to invest in new technologies, then you underestimate the barriers to entry of companies into markets.


Right. Twenty years ago, no one had heard of Google, and Apple was struggling. IBM was still the IT giant. Sony was still the crème de la crème of consumer technology. And what could ever happen to General Motors and Chrysler?



Google didn't innovate. There were a lot of search engines already, they just did it better. They offered free mail? Already available, but google did it better. All Apple did was give people a portable hard disc and call it an iPod, and now that everyone recognized the Apple brand and user friendliness, which existed for a long time, all the Millennial who grew up with the iPod wanted the MacBook Pro and the iMacs and soon the iPad. That's what made Apple grow. Marketing, and Millennials. PCs still dominate business. Apple sells fancy trinkets and phones and leisure computers. The Apple computers maybe are used sometimes in universities and young start up companies and creative industries, but most businesses run windows. General Motors didn't lose to competitor innovation, General Motors and Chrysler lost because they built shittier cars that the Japanese could make better. It's not really that the Japanese had such great innovation, they just had better design.

Competition is important, yes, but competition is not that strong a force to have companies throw themselves in investing millions of dollars into radical new technologies that they haven't found a stable profit maximization model for and has a high chance of a good cost/benefit ratio. AIworld is too radical for the next two decades. Maybe AIoasis, but not AIworld.

Glaucon wrote:
There are a lot of companies that have a monopoly or have few competitors, just because other competitor's aren't big enough or have enough reach in their area.


Sure. British Petrol will probably still be British Petrol, even if they aren't very innovative. But companies like that DO have internal competition. If one branch in one country can safe costs by using particular methods, the overall management will be interested and probably want other branches to do the same thing. This won't always work. But 20 years is a long time, including changes in management, usually.



A single branch doesn't usually have the sheer amount of capital it would take to make a working AIworld a reality.

Glaucon wrote:
Even now, smaller businesses and medium businesses exist in areas simply because the customer is too lazy to economically maximize utility-cost, even though they are selling inferior products at higher prices. People will go to small and medium businesses just because it's close by. Maybe some big companies just don't deliver into that area. And I highly doubt Amazon's 'Delivery by Drone' thing is going to work... it's too easy to shoot things down and steal shit. Maybe even jamming electrical control systems of the drones for 'funsies' as the Iranians did with our CIA drone a few years back. It's not that difficult, and installing countermeasures is costly. That's an example of technology vs working men. UPS will win out, Amazon with their drones won't. At least until they come out drone trucks with ground drone delivery robots.


I don't know about Amazon's 'drones'. But I do know that many local bookshops are struggling, many going under. I usually by my books online, these days, too. Because the selection is much larger, and because the books are cheaper. Record shops? Hah! Sure, people still like going shopping. But competition from online shops is really hurting a lot of retail business. Even for stuff that people generally seem to want to check out in real life. People go to the shoe store and look at what they have, maybe try it on. And then they go online and order it there, because it is 20 dollars cheaper, or because they wanted it in a color that the shoe shop didn't have.



Online stores is basically digitizing stores to what is essentially a mail order service (which existed for many decades) by fancy telegraphing. It's not really new concepts but better technology 'adding on top of it'.

Saying goodbye to manual labor and replacing it all with robots is a bit too over the top in our capabilities, both in the software programming, the mechanical design, and making it economical, all within 20 years.

Glaucon wrote:
My point: Competition as a driving force into investing into the latest technologies as soon as they come up, is overemphasized. Some big corporations, yes, some startups (most startups fail) but not all will plunge into AI technology. Not the entire economy.


The 'entire economy' is rather big. But I think you are wrong. In theory, companies can still operate without using a single computer (for the most part, at least). And yet, nearly every company that is bigger than just one or two people these days does use computers. About 30 years after they became available and affordable for large-scale commercial use. I think that says something about the speed of innovation. Like I said, 20 years is not a short time. I do not buy into the often-told story about how innovation is constantly 'speeding up', necessarily. But I don't think implementation of innovative methods and technologies is slowing down either.



It's one thing to take an existing technology (computers) , and make it better, faster, smaller, smaller component parts, cheaper, and fit it in a room and commercially available...

...and another thing to try to create a shell of an artificial human being that can understand humans and interpret customers, problem solve anomalous situations in out of the box and not preprogrammed ways, and adapt to new situations and tasks. This is after you get them able to walk around and not bump into people.

Glaucon wrote:
Safety is another thing. Anything digital that is connected to the internet can be hacked. It just takes time and resources. But there is not a single encryption that can't be broken with enough time and resources. If you were to connect all public transport to a computer mainframe with automatic systems, with a click of a button you can just have all the buses go and drive off a cliff?


And has that ever stopped anyone? Credit cards are really unsafe (for the banks, anyway) but were and are widely used. Any code can be broken, and yet nearly all of us (as well as companies) are happy to handle something very important (their money through their bank accounts) online, through systems we know aren't completely safe.



Money is one thing. Blood and lives are another.

If your credit card gets charged falsely, the credit card company just cancels the transaction. It's just numbers on a screen, things can always be fixed.

A hacker sending an entire city's worth of trains to derail on the other hand....


Glaucon wrote:And can we really trust an automated bus driver? Well, we KNOW that human bus drivers sometimes drive off cliffs. What if the statistics show that the automated bus drivers get into far fewer accidents? Airplane pilots usually let the autopilot handle most of the flying, and that seems to lead to fewer accidents. An automated system won't get distracted by something, will have a very short reaction-time and (I'll assume) won't drive drunk or fall asleep behind the wheel. I think that this 'fear' of handing over control to a machine is something of a myth when it comes to practical stuff. People seem perfectly alright handling control over to machines when it is convenient.



In the short run automated things may seem safer. In the long run, mass carnage can ensue.

Glaucon wrote:
At least, if you want to do that in todays world, you would have to infiltrate real humans into becoming suicidal bus drivers. But in this new digital world you paint, any shadowy terrorist organization with enough computer resources anywhere in the world can cause mass havoc and mayhem. Digital defense requires a constantly vigilant staff of security experts to monitor the situation. This is expensive. So what happens when the entire economy is linked to technology, an 'Internet of Things', and there's only a few security experts coalesced in a few areas of society, like defense, power plants, a few major city's mass transit systems.... what about the rest? What about smaller cities and their infrastructure?


Eh? Most of what you mention is the case, already. Which is why all these gov. security bigwigs wanting more budget are always going on about cyber terrorism.



I know that is already the case. I'm just saying this persistent problem will continue, and be made even worse if we create an Internet of Things. There will forever be a finite number of cyber security experts, and if we create impossibly more 'zone' for them to cover, there will be a lot of playground for black hat deviants who think they are the masters of the world just because they can write some code onto a keyboard.

Glaucon wrote:
And then, only the second amendment would save us all. And that's assuming judges haven't gotten rid of it yet by creative 'interpreting' of the constitution.


Right. *Cough* Rand Paul for President! :mrgreen:

Can't say I really agree with your analysis. One way or another, in 10 years, in 20 years, in 50 years, a hundred, part of this stuff will come true. Resisting it with a gun left over from the civil war hidden under your grandmother's bed won't stop that. You Americans (yes, I am generalizing) seem to have this romantic fantasy of the future of a dystopia in which you all fall back on pioneer-like individualist 1776-resistance-mode and in which the small community 'fights the power' (and wins). But realistically, in such a struggle, Mat Damon and Neo and all those other resisting types would actually lose, however heroic their fight. If this image of the future has any validity, it will be because people will embrace it willingly. Everyone will want to take the blue pill, instead of the red, to use Hawt's terminology. And those insisting on the red will be labeled crazies and terrorists. And who is to say that the truth of those taking the red pill is the real truth?


Well, maybe to men of 1776, we already are living in a dystopia. Maybe the Unabomber took the red pill. Maybe we all are already taking the blue pill. Who knows? I suppose we are all already used to the current 'enslavement' in today's surveillance society. We should respond with more enslavement of course! :D

You're right. In the end, it is humans themselves who create their golden cages and sit themselves in it. This will all come to pass because people want it. Well, until the country becomes a miniature Elysium movie in 60 years, when people go to their weapons and riot in a vain attempt to push back this juggernaut of digital ascendance, and the elites will just control their little robots and put the 80% of the population back in their place. Robots with guns, this time. Obedient riot police without a conscience, the dream of the corrupt.
User avatar
Glaucon
Posts: 2832
Joined: Tue Nov 03, 2009 9:07 am

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Glaucon » Tue Mar 18, 2014 7:23 am

On the 'It is one thing to... but another thing to...': I didn't say that society would be transformed into this science fiction AI-robot and drone filled fantasy within the next twenty years. I was talking about something more realistic. Something that, potentially, WOULD still pose economic and social issues.

When making predictions on these really big 'whole new world' transformations, I'd like to have a good idea if and when real turing-test proof AI would be created, first. As I see it, that would indeed be that Hall 9000 moment, that singularity. If and when that happens, bring on the droids and think of something new to do for the humans. But I'd not predict that happening within the next 20 years. Unlike some, I do believe it IS quite possible, but it seems the task of creating that shit is really pretty massive. I'd guestimate that it will happen within the next 50 years or so. Maybe it could be done sooner, if we went at it like we went at going to the moon. But I am not sure. I think even the greatest AI experts can't offer more than a very rough educated guess.

Regarding the next 20 years, I was sticking to the 'it is one thing' -side. Not arguing for 'AIworld'. Maybe Hawt was.

Google didn't innovate. There were a lot of search engines already, they just did it better


They just did it better? That IS innovation. But even if it had been Yahoo or Gopher or whatever engine that 'won' the search engine battle, what matters is that something tiny like that can grow into something really big (in economic terms) in 20 years time. You seemed to be arguing that, overall, the landscape of big corporations doesn't change much over such a period. I think it is clear that that landscape did change significantly over the last 20 years. Sure... many mountains are still the same, but some are gone or much smaller, and there are some really big new mountains, now. Same thing with the other examples.

Online stores is basically digitizing stores to what is essentially a mail order service (which existed for many decades) by fancy telegraphing. It's not really new concepts but better technology 'adding on top of it'.


It doesn't matter that it is nothing 'new' really (though, it actually is, because ordering things from an annual catalogue is different from ordering them from many different websites with changing content), what matters is that that sort of a business didn't impact retain much, before. It does now.

Saying goodbye to manual labor and replacing it all with robots is a bit too over the top in our capabilities, both in the software programming, the mechanical design, and making it economical, all within 20 years.


Again, I never said any of that would happen in the next 20 years. Maybe Hawt thinks it will happen that fast (though he didn't say that either), but I do not. I do suspect that it will happen at some point, though, unless modern civilization 'collapses' or something before we get there. You might even live to see some of it (I probably won't). Or maybe you won't. We won't really KNOW how hard it really is to create full human-like AI until we made it.

I'd say that there are roughly three 'paths' that people point at as the most direct route to creating substantive human-like AI:

- The 'Good old Fashioned' approach (GOFAI) in which people try to use what computers are good at in an attempt to make them perform tasks that seems to require thinking, by logical symbol manipulation, essentially. Like Asimov's robots, or Mr. Data. Usually, that means giving them access to a huge data-base and lots of calculation power. Like the Jeopardy playing computer. Or google translate (though that uses 'statistical info' as well). This produces the most impressive results (it is the method used to power those chatbots), generally, but it clearly doesn't work in any way that resembles how we think.

- The connectionist approach: This imagines that human thought isn't a matter of deductive reasoning or symbol manipulation but of (spontaneously) emerging patterns within a system of connected units. In a sense, it seems to run with another idea of what the central elements of human mental processes are and it tries to reverse engineer those processes to try and create a system that can 'learn' how to perceive and thing itself. Like those German guys were doing with little Asimo. This stuff is interesting, but... as far as I am aware, I don't think that this sort of approach really leads to much that is useful outside of research, just yet.

- The 'just copy the brain' approach: This one is pretty straight forward: The idea is simple: If we can understand the human brain, we can model it inside a computer. So, some of the projects in this field involve making very very detailed 3D maps of actual working brains and the neural connections and the activity inside of them and rebuilding all that as a working virtual model. In theory, that way, you could recreate human intelligence without really needing to understand how it works, even. Personally, I believe this is a rather tall order, just yet.

The second approach seems to be the most intelligent one, to me. One of the reasons why I think AI hasn't gone as far as it might have is because it seems that the second approach doesn't have enough immediate commercial spin-off. Baby robots that are learning may be cute, but you can't really expect them to 'do' much. So, money pumped into it doesn't lead to any return on investment (unless that sort of technology really starts making big leaps).

Eventually, one way or another, they will figure it out, though. And when that happens, all bets are off, and you will get the AIworld you are talking about.

I know that is already the case. I'm just saying this persistent problem will continue, and be made even worse if we create an Internet of Things. There will forever be a finite number of cyber security experts, and if we create impossibly more 'zone' for them to cover, there will be a lot of playground for black hat deviants who think they are the masters of the world just because they can write some code onto a keyboard.


Those dangers are just part of scientific and technological development. Guy Falkes and friends tried to blow up Parliament using gun-powder, because that stuff was the most potent thing available to him at that time. In this day and age, his modern day equivalent might be trying to construct a nuclear weapon in his basement (fortunately, you need quite a few hard-to-get parts for that) or a chemical weapon or something. Luckily, most crazies trying stuff like that aren't smart enough to manage any of that. Instead, they make simple bombs (like those Boston bombers) or just use the pretty deadly technology readily available to people (guns, cars or, if they can manage it, planes). I think the same is generally true (or going to be true) for those that want to wreck cyber-havoc. It isn't a 'fight' between a few cyber security experts and an army of hackers. It is one or would be one between 99+% percent of the people versus a few total nut-jobs and a somewhat larger group of potential criminals out for monetary gain.

Well, maybe to men of 1776, we already are living in a dystopia. Maybe the Unabomber took the red pill. Maybe we all are already taking the blue pill. Who knows? I suppose we are all already used to the current 'enslavement' in today's surveillance society. We should respond with more enslavement of course! :D

You're right. In the end, it is humans themselves who create their golden cages and sit themselves in it. This will all come to pass because people want it. Well, until the country becomes a miniature Elysium movie in 60 years, when people go to their weapons and riot in a vain attempt to push back this juggernaut of digital ascendance, and the elites will just control their little robots and put the 80% of the population back in their place. Robots with guns, this time. Obedient riot police without a conscience, the dream of the corrupt.


I am not sure this AIworld future would be all that dystopian. Maybe it won't be a world of evil droids with guns firing on the masses. Our technological advances made totalitarian 1984-like states like the regime in East Germany (and present-day North Korea) possible, but if we live in our own 1984 world with Big Brother NSA watching, our certainly SEEM a lot more free than the lives of the people in Orwell's version.

We do live in democracies. In theory, people have the ability to select leaders that look out for their interests. In theory, the people have the power of self-determination and a say in what their world should be like. I know that generation X-ers and gen Y2K don't believe in old fashioned stuff like having power through collective action and politics on a big scale, generally, but it ought to be possible, in theory at least, for the people of a country to make such new technology work for them, rather than against them.

If you have time, perhaps you should pick up one of Ian M. Banks culture novels at some point. Space Operas, those, but they do depict a sort of Asimov-like society in which AI has effectively superseded human intelligence (and taken control without a fight), with humans living generally very safe lives alongside their much more powerful children as a kind of beloved and cherished pets. Maybe that will over a nicely contrasting vision to the Elysium-style dystopias.
User avatar
Theoden
Posts: 398
Joined: Wed Oct 05, 2011 6:20 am
SL Name: Theoden
Caste: Warrior

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Theoden » Tue Mar 18, 2014 11:19 am

Prime Bbcode Spoiler Show Prime Bbcode Spoiler:
Glaucon wrote:On the 'It is one thing to... but another thing to...': I didn't say that society would be transformed into this science fiction AI-robot and drone filled fantasy within the next twenty years. I was talking about something more realistic. Something that, potentially, WOULD still pose economic and social issues.

When making predictions on these really big 'whole new world' transformations, I'd like to have a good idea if and when real turing-test proof AI would be created, first. As I see it, that would indeed be that Hall 9000 moment, that singularity. If and when that happens, bring on the droids and think of something new to do for the humans. But I'd not predict that happening within the next 20 years. Unlike some, I do believe it IS quite possible, but it seems the task of creating that shit is really pretty massive. I'd guestimate that it will happen within the next 50 years or so. Maybe it could be done sooner, if we went at it like we went at going to the moon. But I am not sure. I think even the greatest AI experts can't offer more than a very rough educated guess.

Regarding the next 20 years, I was sticking to the 'it is one thing' -side. Not arguing for 'AIworld'. Maybe Hawt was.


Then I guess we agree on the same things then

Prime Bbcode Spoiler Show Prime Bbcode Spoiler:
Google didn't innovate. There were a lot of search engines already, they just did it better


They just did it better? That IS innovation. But even if it had been Yahoo or Gopher or whatever engine that 'won' the search engine battle, what matters is that something tiny like that can grow into something really big (in economic terms) in 20 years time. You seemed to be arguing that, overall, the landscape of big corporations doesn't change much over such a period. I think it is clear that that landscape did change significantly over the last 20 years. Sure... many mountains are still the same, but some are gone or much smaller, and there are some really big new mountains, now. Same thing with the other examples.


Prime Bbcode Spoiler Show Prime Bbcode Spoiler:
Online stores is basically digitizing stores to what is essentially a mail order service (which existed for many decades) by fancy telegraphing. It's not really new concepts but better technology 'adding on top of it'.


It doesn't matter that it is nothing 'new' really (though, it actually is, because ordering things from an annual catalogue is different from ordering them from many different websites with changing content), what matters is that that sort of a business didn't impact retain much, before. It does now.


in·no·va·tion
noun
: a new idea, device, or method
: the act or process of introducing new ideas, devices, or methods

I thought innovation means it has to be something completely new? Like an invention. I suppose it doesn’t have to be. I googled more and it seems innovation with relation to technology means just improvements on something already existing.

innovation
noun (Concise Encyclopedia)
In technology, an improvement to something already existing. Distinguishing an element of novelty in an invention remains a concern of patent law.


Ah well. Semantics. Anyway, I wasn’t arguing that the landscape of corporations don’t change, I was arguing that truly new radical technologies didn’t cause that change for the most part. But if you’re talking about improvements on already existing technologies then you would be right, per the encyclopedia technology definition.

Glaucon wrote:Baby robots that are learning may be cute, but you can't really expect them to 'do' much. So, money pumped into it doesn't lead to any return on investment (unless that sort of technology really starts making big leaps).
I guess they can make more intelligent Furbies or something :P

Glaucon wrote:
I know that is already the case. I'm just saying this persistent problem will continue, and be made even worse if we create an Internet of Things. There will forever be a finite number of cyber security experts, and if we create impossibly more 'zone' for them to cover, there will be a lot of playground for black hat deviants who think they are the masters of the world just because they can write some code onto a keyboard.


Those dangers are just part of scientific and technological development. Guy Falkes and friends tried to blow up Parliament using gun-powder, because that stuff was the most potent thing available to him at that time. In this day and age, his modern day equivalent might be trying to construct a nuclear weapon in his basement (fortunately, you need quite a few hard-to-get parts for that) or a chemical weapon or something. Luckily, most crazies trying stuff like that aren't smart enough to manage any of that. Instead, they make simple bombs (like those Boston bombers) or just use the pretty deadly technology readily available to people (guns, cars or, if they can manage it, planes). I think the same is generally true (or going to be true) for those that want to wreck cyber-havoc. It isn't a 'fight' between a few cyber security experts and an army of hackers. It is one or would be one between 99+% percent of the people versus a few total nut-jobs and a somewhat larger group of potential criminals out for monetary gain.



The thing is, the dangers of technological advancement in those cases are easy to counter, since they require physical materials and are restricted by physical logistics. Materials for chemical weapons and nukes, for example. Even the Oklahoma city bomber could only take out a single building.

The dangers with this new era of interconnectedness means it is a lot easier for a few to take over… and when they do, since everything is interconnected and centralized, the power they have is exponential. No more just Parliament. No more just a government building. No more just a marathon. You can have catastrophes occur throughout the entire country by causing automation malfunctions. Automated Locks, automated sprinkler systems? Lock people in their houses, drown them. Automated planes become giant missiles, train derailment, cars ploughing into huge crowds. Automated chemical waste disposal placing it into residential areas.

How is it a fight between 99% of people versus a few nutjobs and criminals? 99% of people aren’t technology and security experts. It will be just a few tens of thousands of security experts spread in crucial sectors of society like making sure nuclear power plants don’t blow themselves up, and the rest of society will have to fend for itself and hope the generic anti virus they buy will hold them at bay.

Anonymity. The ability of a few to have exponential power in the shadows. This isn’t just going to be a bunch of low tech nutjobs and criminals. This is going to be well funded cyber terrorists who want to see the world burn or nation state supported shadow proxy ultranationalist cyber militias who want to take down the West. “Oh shit the entire transport infrastructure on the eastern seaboard is fucked for a few days and the Dow Jones plummets!” “Wasn’t us.” Says other countries. Right now they do idiotic things like take down NATO websites because of the Crimean situation. Which is the equivalent of taking down a poster and more vandalism than a real security breach. But once the power of the hacker against an increasingly interconnected Internet of Things come around, powerful entities will pour more money into shadowy hacking groups that are more professional and more dangerous. Not these current low budget nerds who can only run DDOS attacks with a botnet.

Glaucon wrote:

I am not sure this AIworld future would be all that dystopian. Maybe it won't be a world of evil droids with guns firing on the masses. Our technological advances made totalitarian 1984-like states like the regime in East Germany (and present-day North Korea) possible, but if we live in our own 1984 world with Big Brother NSA watching, our certainly SEEM a lot more free than the lives of the people in Orwell's version.

We do live in democracies. In theory, people have the ability to select leaders that look out for their interests. In theory, the people have the power of self-determination and a say in what their world should be like. I know that generation X-ers and gen Y2K don't believe in old fashioned stuff like having power through collective action and politics on a big scale, generally, but it ought to be possible, in theory at least, for the people of a country to make such new technology work for them, rather than against them.


People look at things in the short run and don’t care about the long run sufficiently. In a more interconnected technological world, exponential catastrophes will occur through third parties. It is inevitable. And when it does, Orwell’s vision will come true when people willingly give up their freedom for security. All it takes is one catastrophe and one election.

Glaucon wrote:
If you have time, perhaps you should pick up one of Ian M. Banks culture novels at some point. Space Operas, those, but they do depict a sort of Asimov-like society in which AI has effectively superseded human intelligence (and taken control without a fight), with humans living generally very safe lives alongside their much more powerful children as a kind of beloved and cherished pets. Maybe that will over a nicely contrasting vision to the Elysium-style dystopias.


Humans living as cherished and beloved pets, with an artificial super-intelligence-overmind to tell them what to do? Effectively becoming their God?
Man was not made to be caged or become a pet of robots. How is that not a dystopia LOL. That’s like the rationale of every A.I. takeover of humans in literature and cinema. “You humans are so like children. We must take over to make sure you don’t make war anymore. We can solve all your social ills… just do as we say. Trust the overmind.” Seriously, watch ‘I, Robot’ lol.
User avatar
Glaucon
Posts: 2832
Joined: Tue Nov 03, 2009 9:07 am

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Glaucon » Tue Mar 18, 2014 1:50 pm

Theoden wrote:Humans living as cherished and beloved pets, with an artificial super-intelligence-overmind to tell them what to do? Effectively becoming their God?
Man was not made to be caged or become a pet of robots. How is that not a dystopia LOL. That’s like the rationale of every A.I. takeover of humans in literature and cinema. “You humans are so like children. We must take over to make sure you don’t make war anymore. We can solve all your social ills… just do as we say. Trust the overmind.” Seriously, watch ‘I, Robot’ lol.


Well, if you have time, pick up one of those books. In them, lots of people don't like the idea either, especially those from more 'primitive' cultures that think in terms of power, etc. In that universe, the AI controlled civilization tends to be quite a bit more powerful than any of the non AI controlled people, and for the most part, they are careful to let humans retain the illusion of 'mattering'. ;)

Science fiction, of course. Then again... we live in a pretty big world ourselves, in which few of us really have power over much beyond ourselves and those around us, or much by way of influence. Maybe we 'feel' we matter more than we actually do, already.
Hawt Sommer

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Hawt Sommer » Wed Mar 19, 2014 1:33 pm

U.S. Navy 'stealth drone' takes to the sea for tests: The autonomous X-47B is hoped to be first carrier-borne unmanned aircraft.

This stealth drone set to be the world's first unmanned, robot aircraft piloted by artificial intelligence rather than a remote human operator has taken to the sea for tests.

Read more: http://www.dailymail.co.uk/sciencetech/ ... z2wQjXaIx5


Currently Drones make up 1/3 of the Airforce's aircraft.

And speaking of pets...

Meet Big Dog (Five Years Ago)
http://www.youtube.com/watch?v=cNZPRsrwumQ

Meet Atlas (2013)

http://www.youtube.com/watch?v=zkBnFPBV3f0

From four legs to two legs in five years... and.. kind of looks like a terminator robot.. Gah!!!! :shock:

In December 2013, Google Acquired Boston dynamics along with 8 other robotics companies.
User avatar
Glaucon
Posts: 2832
Joined: Tue Nov 03, 2009 9:07 am

Re: "Software (Bot) Substituion" A Gorean Forums Think-tank

Postby Glaucon » Wed Mar 19, 2014 2:42 pm

Yes, a robot that can walk around like 3CPO isn't really far off. Robotics really is moving along. Doesn't mean they can talk or think, and there are huge issues when it comes to weight and energy supply (you can see that that Atlas robot has lots of cables connected to it), but I do think that we will see robots that can pretty much walk around like humans or animals soon enough. A lot of the technologies needed to make that happen (small but powerful power supplies, tiny mechanical units, and so forth) are being improved because we want that stuff for other technology as well.

Military applications are obvious (we already have the robots that are used to deal with explosives and mines and the areal drones, of course). The small vacuum droids are becoming more popular. I'd guess that it will be a while before there will be mobile robots freely moving about in public, though. Those self-driving cars would be the first of that sort, I guess.

Return to “News”

Who is online

Users browsing this forum: No registered users and 1 guest

cron