And now for something completely different...

Toctoc, Kadi here.

Nope, again no new Fluffiness. Sorry. Today we're going to save mankind! And for once, I'm mostly serious.

 

About two months ago, I was linked a little Youtube video titled "Humans need no apply". Some of you may know it already, some may not. It's on the history and more importantly the future human (and horse) labor and employment. When I watched it for the first time, I was stunned out of my mind. Its contents are something that everybody in our generation should have seen, because it's it's coming and chances are, we'll have to deal with it. And so far I had no idea it was coming.

The link is here, the contents speak for themselves.

I don't actually want to start much of a political discussion here. Lightly discussing Justice and Fluffiness, Darkness and Dawn suits us just fine.

I don't actually want you to do much. For now, I doubt there's much that can be done.

I don't even want you to agree with the contents, or me. If anything, I hope you can convince me it's thorough and utter bullshit.

I just want you to know. I want people to know. So please, watch the video, take in what it says. Show your friends. Show your families. If you know some appropriate people, show your teachers or lecturers, maybe, colleagues or bosses if your relationships are right. Post it in your favorite forums, if it's not there already. And while hypocritical me doesn't use them: Put it on your twitters and facebooks.

This is "need-to-know", and not the kind where most people must not know.

Thank you.

 

——Kadi

Leave a Reply

Your email address will not be published. Required fields are marked *

97 thoughts on “And now for something completely different...

Open Message Board
  1. anonYmous

    Yeah, after watching the vid it reminds me of my condition..
    at least when that happened I already died.. Hopefully..
    I'm 21 jobless btw.. >.>, in my final year of college and have no experience of working in society..
    Please pray for my well being..

    Reply
  2. anonymus.suika

    This really reminds me of this short story, http://marshallbrain.com/manna1.htm
    We might be able to reach the next step of the civilization but the road there is really bumpy.
    There will be a lot of people get hit by the turbulence it will not end well for them.

    Reply
  3. DJ

    If you produce something, you have to try and minimize time and cost to produce it and try to balance the price you are asking with the quantity you are selling. That has always been the case with capitalism. To look at robotics and only see that a robot takes away a job from a human is the wrong way to look at it. Nothing is ever black and white. If the company would rather pay more money and hire people, then it will loose against its competition who use robotics and the end result is that all people working at that company are out of jobs. Who goes to the local supermarket and, given the choice or two similar products, takes the more expensive one? I certainly don't.

    This process has begun with the industrial revolution in the 18th century. Imagine how many jobs were lost when they first stopped using completely handmade crafts and started switching to the manufactured way to produce? Every village once had a blacksmith. Now, not even the big cities have any. Isn't it terrible? But wait. I'm driving a car. That would have cost me a livetime of savings with all handmade work. Now i can affort it for the savings of months, not years. Advantages of capitalism.

    Robotics can only do so much. If a process is repeating and does not need a human to think, it can be automated. If it includes making creative decisions then it will always require a human. No robot ever can be creative, can react to a surprise that hasn't been forseen. No book can ever be written without a human. No building designed. You are not really saying that robotics made you afraid of no longer being needed, right? Even the most stupid human can always do better than a robot! Not that i'm telling you that you are stupid, mind. :-)

    Other points are also well made. I'm rather surprised google tries to get cars to drive themselves. Accidents will happen and then people will sue google, no way around it. After all, google has money and if something goes wrong, supposedly they are to blame for it. One can hardly sue a robot.

    There is so much more stuff that humanity as a whole must pay attention to urgently. Have you ever seen the Home (2009)? We are destroying our life habitat. Or perhaps we want to take a look at the money spend to make war (huuuge) compared to that spend to get our hands at the huge resources that are just waiting in space (very small)?

    Change is always hard. No matter that nature dictates change to survive.

    Reply
  4. Stranger Danger

    Look the main point is bots are coming end of story. I studied mechatronics and see it coming a long time ago unlike my dad who forced me into a business degree. There are things they can and cannot do even with learning capabilities.

    They make errors specified within the programming/fault of programmers. There are jobs like being the judge that cannot be replaced as they require a human "heart" or soul but many repetative and compiling jobs will dissapear. They can even teach themselves to program each other but like some of you said they cannot learn from very unexpected things or make super spontaneous decission like our cavemans did.

    They don't take higher risk but they take safe decissions. After playing many card games i can tell u taking safe decission and making no wastage can win you the game 85% of the time. But i take very different safe decission that seems wasting a lot more resources to others and win those who takes the safe decission almost all the time. This due to risk taking and believe in the unknown and adapting to situations. Something an ai cant be programmed to do all at the same time cause they will output crazy answers.

    Its a logic based thinking so they can't output it especially my moves cause i just make "stupid" moves that would cause losses if it fails... And yes there are things like soul to certain things cause its the easiest way to explain sixth sense for certain people who can feel the fourth dimension on things Like telling apart the music. When i heard the music in the video i could tell immedieately it was weird cause there was no soul... Its beautiful but souless as if no human expression or heart.

    Art is an expression thing and there are people who truly can feel these things but majority of people who claim to be able to can't and gets fooled by the machines so its possible to replace this sector with bots as well. Its something you need to experience or be born with and be talented in it. Thus replaceable with bots due to lack of these people and the general populace can't tell the difference. Btw i play the trumpet so yea had to learn it. There are subtle sounds differences that can form different sounds in each notes to form expression and a person's soul even with the same tuning. Thats why perfect sound 100% is souless as if you are reading a book to children in monotone. Teach it to put in these subtle differences and same problem it will go crazy or not play well at all.

    This does mean that the majority of the jobs can be replaced but there are certain thing that can't be replaced but as the video said as long as most of them can be replaced its good enough but i say that the parts that can truly be replaced are around 38% as compared to the video because we still need people who know these thing to help program the bots in he first place to find more faults so as to make lesser errors more quickly rather than having them teach themselves after making the error themselves.

    Everything is affected but it just means that society will have to change their value system if this were to occour. But it would be hard for this as we have depended on our current system for many hundred thousands of years and maybe only a couple of thousands for certain countries. But there is massive changes coming over the horrizon and its uncontrollable and soon even the alternative energy documents that were being witheld by universities due to investors would finally be brought to light if this were come to pass and there will at least be benefits and downsides to all this.

    Sorry if i confused you or anything but i just woke up and yeah... English is my second language. And btw the video was biased if you did not noticed caused they covered the faults and what the bots can't truly do. But i agree that they are coming and can shake the world as we know it but not at the scale he was saying it at but at probably around 3/5 his scale and there are some people who wants to fight this but this will probably end in tears unless the general populace threatens their government. Btw 3/5 his scale is bad enough if you did not notice.

    Reply
  5. All Night

    The sheer walls of text on here :o. Anywho, I work a job which can't be replaced by a machine, at the very least for the next 30 or so years..

    Reply
  6. bladerain

    This idea occurring in the future is quite possible. However, there are limits to how many jobs can be replaced. This will make the human race lazy at one point. Although some professions may be unaffected, this will cause a lot of poverty and other social problems.

    Reply
  7. DedWards

    I haven't watched the video yet, but I get a general idea of what it's about from the comments here. I'd like to give my view on robots taking away jobs.

    I live in a 3rd world country (South Africa) and lack of jobs is a constant problem here. Although I agree that the amount of jobs that robots can take over are limited, the problem remains that those same jobs could be given to humans, thus lowering poverty, crime, etc. So the biggest problem I see is middle to lower class income people could be forced to resort to crime just to survive because robots reduce the number of jobs available.

    This will hit 3rd world countries worst, but I can see it having an affect on 1st world countries too.

    Reply
  8. colddfire

    Well... after watching the video, the only thing that cross my mind was: IT'S SKYNET!!
    ....nuff said

    Reply
  9. Eudaimonion

    I think Asimov is a good source of information about this subject. As he explains, when robots get more complicated, we have to keep pace with them to repair them, until we hit singularity. After that, it's fine to move to a hedonistic society where really, all humans do is play games, socialize and view/create media all day. It wouldn't be unheard of for us all to gain game compulsion. It's not a bad future, despite what people might think. And, yes, since robots can replace professors, we can all become highly educated and even help them solve things. Or just have generally interesting media with actual scientific basis. Who knows.

    Oh, and I noticed a common thread among comments was 'humans don't trust computers'. This is a lie. If it weren't so, we wouldn't cry when information was lost-- we wouldn't have the information on a single computer in the first place. The problem is us trusting them too much.

    Reply
    1. Sonoda YukiSonoda Yuki

      It's not that we don't trust Computers, it's that we don't trust computers that make obvious mistakes even when they are more effective than humans in the long run. Most of this is because of a lac of indicators of reliability. How certain is Apple maps that it isn't telling you to drive off a cliff? It may not be very sure at all, but if it doesn't tell you that it's instructions may be unreliable, and it fails you anyway, you're going to despise it for the (short) remainder of your life.

      Reply
      1. Reaper Phoenix

        Computers make obvious mistakes because we can't program common sense into it, YET. There's a novel about it by James P. Hogan called The Two Faces of Tomorrow where they make super intelligent AI, and how they solve the no common sense problem. AND THE COST.

        Reply
        1. Eudaimonion

          Actually, as a programmer, I can tell you we can create common sense in machines-- even morality. We call this 'behavioral heuristics'. We can have a machine assess the variables present, rank the actions it needs to do, cut away branches that would potentially hurt another or fail the process, and have it analyze them all. This is what you would call 'common sense' and often 'morality'.

          Reply
          1. DJ

            It is still a conditional response based on arbitrary values given by its programmer. It emulates common sense in an expected and predicted environment. The true decision how to react is still done by the programmer and not the machine itself.

  10. Forumreader

    As engineering student this whole premise to put it simply is crazy. I had a conversation about this with my brothers girlfriend in the car the other day in which she brought up a private sector job the lifeguard. As she has that job as a part timer she was really concerned that a bot might one day take over her job and I tried to convey to her that the technology shes thinking of is at least 150 years into the future. Machines for the most part are limited by all the situations we prepare them for. Going back to the lifeguard consider for a minute what might happen if someone is drowning. How did that person start drowning? Well there is about a billion different factors that could have been involved (obviously and overstatement but I am trying to impress Murphys law upon you). Now an engineer could probably make a machine that could respond to 10 different situations for a relatively cheap price. The rest of the situations are ignored and buying a robot would be pointless since you would still need a human for the rest of the situations. Or you could make a specialized machine that can respond to each of those problems but, that would take about 10million dollars (at minimum) would be twice the size of the entire facility and would need to be constantly maintained. So think of your price range being around 20 million to start and over 10k a year on maintenance and 15 years from now you need to revamp you system (costing you another 5mil).

    Now lets go further and say that one day we program ourselves completely out of work (something that I would consider impossible for at least 1 thousand years).... whats the problem? If the entire process of finding supplies, collecting it, transporting it, and refining it into goods is managed by machines and not a single human is employed you have created a utopia where all costs of human living are carried by the environment. Think about it, currently when you pay for something you are basically trading work. You take the work that you have converted into a tradeable form (money) and use it to buy somebody else work. Even if a car making process is 90% automated consider all the people involved in finding the materials for the car and supplying the energy for the automation system. If a machine could be made that could do all that on its own there would be no value to anything as no people are doing any work. Now lets consider when 45% of the populous is unemployed. By the way, the technology that could keep 45% of the populous from working is huge majority of jobs and all it would take is one clever engineer to take that system and alter it to eliminate everything but the creation jobs (designing more of these machines, solving mysterious of the universe ect). In that case since only work is being done by these engineers everything would be absolutely dirt cheap and you could buy a maid robot to take care of you for the rest of your life for peanuts. The bigger problem in that case is everyone would be unmotivated to do anything (including those engineers) and since you cant trade anything anymore (since nothing has value) housing would be a nightmare (at the same time it wouldn't because if you had the technology to eliminate all work then colonizing a planet would be a piece of cake and then you have people that either travel in space or finding a nice piece of nowhere to settle down).

    In conclusion, at the moment designing everything out except very very VERY basic jobs costs more than it would make in the long run and if you COULD program out a huge segment of the populace it would only very briefly hurt economically because the work saved by designing them out would make life oh so much cheaper.

    Reply
    1. AoriiAorii

      "Well there is about a billion different factors that could have been involved"
      Yes and a machine can either be programmed or self-taught to deal with all eight billion different factor (like the finance companies' auto-traders; they scour the internet new sites and social network postings for 'hints' about the economy then trade based on it). Software AI technology is far more advanced than most people realize. Sure, it might take years and millions to set it up. But once it's up, it replaces hundreds of jobs forever.

      "So think of your price range being around 20 million to start and over 10k a year on maintenance"
      You should check out how expensive the first computers were to build and maintain... and now? I mean come on, you're engineering. Most engineers don't invent anything new, just making existing tools and processes more efficient. Multiply that by millions of engineers and decades of time and you've got...

      The concept of "permanent vacation" is well established in scifi, where the people of any sufficiently advanced civilization works because they want to, not because they have to. The problem is the transition. Failed social transitions leads to periods of unrest and anarchy that kills millions, billions.
      Just think of the French Revolution or the October Revolution. Don't think it's limited to politics.

      Reply
      1. Dedecou

        And if a Machine is 'self taught' when saving drowning people(I mean, the Machine is beginning to learn when it's already saving those people)?

        'Yes and a machine can either be programmed or self-taught to deal with all eight billion different factor (like the finance companies' auto-traders; they scour the internet new sites and social network postings for 'hints' about the economy then trade based on it). Software AI technology is far more advanced than most people realize. Sure, it might take years and millions to set it up. But once it's up, it replaces hundreds of jobs forever.'

        And many of those factors are much harder than just *searching the internet*, for example the lifeguard needs to detect patterns that let the machine know when the person is drowning or not. Of course, though for humans it's easy, for machines it would be lot harder.

        'You should check out how expensive the first computers were to build and maintain... and now? I mean come on, you're engineering. Most engineers don't invent anything new, just making existing tools and processes more efficient. Multiply that by millions of engineers and decades of time and you've got...'

        It seems like you are considering on creating a facility with PCs. But for such task, you would need to have powerful servers that needs a structure of their own,a maintenance crew on standby(considering it's a rather important facility).
        Not to mention that the 20 million figure that he gave, is more than to buy just the computers...
        And of course, to add to the price, you need to the design cost, as those *millions* of engineers don't work for free and of course, they need computers of their own to run their calculations and space to run test's.
        It's much harder than *hire a million of engineers* and they will design and build it.

        Engineers are also the one's who invent pretty much everything.
        Scientist do research, Engineers apply it.

        'The concept of "permanent vacation" is well established in scifi, where the people of any sufficiently advanced civilization works because they want to, not because they have to. The problem is the transition. Failed social transitions leads to periods of unrest and anarchy that kills millions, billions.'

        That is why it's science FICTION.

        Also not to forget to mention that modern computers will evolve till a point where it can't anymore. Miniaturization of transistor per example would eventually reach the limits at the atomic level.

        Reply
      2. Forumreader

        You fail too see the point, yes those technologies exist but as I said they cost millions and if you could make a system that DIDN'T cost the next 100 years of earnings of the company you will have invented a technology that can do anything. Think about it a program that self learns and can deal with all 8 billion problems if it can literally be converted to do ANYTHING. You are imagining a Baxter, Something that while currently inefficient is self learning and can do lots of things. Unfortunately Baxter will never be able to handle all problems why? Because he is STATIONARY and can only do things within his reach thus even if he can learn at the 100 times the speed of a person he will NEVER be able to fix all 8 billion problems. A Baxter who can do all 8 billions things is basically a machine on the level of the terminator.

        Reply
        1. krytykkrytyk

          However, you're missing one thing. Economics doesn't need multi-purpose robots, but specialized robots who can intelligently respond to changing situations. Take for example the government office, imagine that instead of a lady in chair in the office, there's self-service with just a few buttons, and it accepts your petition automatically. How many government workers all over the world will lose their jobs? And it's not really a high sci-fi either. You can already have your legal bussiness done through internets right?

          The other point, this video doesn't promote "fighting" against the rise of the machines, instead it promotes thinking and countermeasures. However, not countermeasures against losing jobs, but how will society change in the future because of the technology taking over the jobs. Currently if you don't work you don't eat, the society and every country is functioning on labour. However, what if in 50 years humans will no longer have to work? What if all work will be done by machines? What of the labour, money? Will people be sitting on their butts and playing games all day long, while the robots will do everything for them?

          Reply
          1. Forumreader

            Those robots that can respond intelligently to changing situations don't exist yet (for the most part). Sure you CAN make a robot that responds to 1/4 the situations but even then robots have very clearly defined limits. My point with multi-purpose robots is that is what would be needed to cause the huge amounts of unemployment your thinking of. Second a specialized robot who can intelligently respond like the car can only do so much and has to be super specialized. Not only that but the car will only be able to replace very very VERY basic level jobs (like taxi drivers). If you go one step up the driving ladder and try to remove truckers your system becomes a quite complicated since while you would be using the same basic technology as the self-driving car you would need to make it a whole hellva lot more complicated to deal with all the issues a trucker could face. Finally a system that can intelligently respond to changing situations currently requires a super-computer and if that lady's job is something that you can do with self service by god I want it replaced right now. Anything that only requires a screen and some buttons to select what you what is a job that even if you have convinced yourself I will never see the benefit of and even if they say they need that job for their living I have seen people in far tougher situations with far more liabilities who managed to find ways to live. My main point with this is any job low-skilled enough to be able to be viably replaced at the moment should and as we always have people will find a way to deal with it and any machine capable of replacing a higher tier job currently is so far from being viable it isn't even funny.

            Since I have made it pretty clear I am in favor of letting the machines replacing jobs to answer your second paragraph I suggest you read my original posts 2nd paragraph. I make it pretty clear that as machines eliminate jobs everything becomes cheaper since you are removing work from the system and putting the cost of that work on the environment (energy costs, waste, ect) and that if cost is carried completely by the environment then nothing currently (other than land) will have any value.

  11. Dedecou

    I didn't watch the whole video, but...
    It's a problem more for you American(and Europeans) though(if it's a problem at all), in many countries labor is very very cheap and it wouldn't be cost effective to replace it with a machine. And of course, the technology isn't that advanced at this point.

    Reply
    1. AoriiAorii

      ...said the camels of Arabia to European Horses.
      The same came to them but a few decades late when truck prices dropped enough.

      Reply
  12. Xero

    Local government has the responsibility to protect the national interest. Yes, multinational companies with deep investments can destroy the local economy when such technologies are actually wide spread enough but it's not so much as the technology itself that will cause the most damage but the policies regarding them.

    Reply
    1. AoriiAorii

      "Local government has the responsibility to protect the national interest."
      History is filled with self-destructive governments who did everything wrong because they were managed by short-sighted idiots.
      And modern 'Democracy' is short-sighted by nature, because politicians don't need to care about two decades into the future when they've already retired in luxury and wealth.
      The US is already a freaking plutocracy. Who will follow?

      Reply
      1. KadiKadi Post author

        The first chancellor of "modern germany", Konrad Adenauer, was told the German welfare system he was about to start was flawed and would collapse in 50 years. His reply, as it's told? "Why do I care, I don't need to win an election then." And hey, the system is on its knees today!

        Reply
        1. AoriiAorii

          ...You just broke every good thought about Konrad Adenauer I ever had.
          Why, Kadi, why must you ruin me so!????
          (lol just kidding).

          Reply
      2. Reaper Phoenix

        One of the weakness of democracy is that it's a popularity contest. Most politician in a democracy are only concern for short term gains or the popularity of the decision they're making. Unfortunately the right decisions aren't always popular or profitable in the short term.
        An ideal democracy can only happen if the voters are very well educated and WISE to 'know' and want what is good for them in the long run. Very unlikely.

        Reply
        1. Anon

          "Vote: The instrument and symbol of a freeman's power to make a fool of himself and a wreck of his country."

          Reply
        2. bladerain

          I agree with that. If the majority is foolish and will head towards destruction, the government follows soon.

          Reply
  13. Lu Pu

    I can totally see this coming, but meh, somehow or the other society will deal with it.
    Not necessarily in a good way, but it will handle it somehow.

    Reply
  14. Owl

    lol, now that this is over with, let us concentrate on the really important problems.

    Like after this translation, we won't be seeing Yun for a while! *cry*

    "I want Yun-nee san!!!" *throws tantrum*

    Reply
    1. Owl

      Just to clarify, I'm not talking about the speed of translations, but the speed of publication of the originals. Oh well. Such is life.

      Yun!!! *cry*

      Reply
  15. Svelte

    Speaking as a computer engineer the video is accurate in some ways, but mostly opinion rather than fact. BTW that's the TLDR if you're scared of textwallgasms. It is true that we programmers can write code for machines which mechanical engineers and electrical engineers will put together to do manual tasks like coffee making. It is true that we can write programs to trade stocks or diagnose diseases which can have a form of 'self-learning'. But if I have one complaint with this video it's that it doesn't really take a close look at these things.

    Take 'self-learning' for instance. The basic premise, as I'm currently aware of it and while I know an automation engineer and talk with him frequently I am not one myself, of modern day self learning machines is to follow a four step program.

    Step 1) Gather data via sources like the internet, through observation, or through humans via data dumps - this step basically continues beneath the surface during all other steps btw
    Step 2) Begin making predictions once a certain data threshold is reached - generally these will just be predictions, a stock trader bot at this point wouldn't be trading stocks yet, just predicting how they will act
    Step 3) Compare predictions to results in order to tweak data
    Step 4) Once the ratio of accurate to inaccurate predictions reaches a certain threshold begin 'work' that is to say at this point the stock trading bot would begin actually trading stocks.

    This is something that can be incredibly complex, especially in cases of multiple-purpose bots like Siri, but it is very doable currently.

    Original thought, currently, is not so doable as far as I'm aware and shouldn't be in the foreseeable future. Though who knows, maybe there's a robot sitting in the White House basement right now, writing Obama's speeches. But back to the point, take the music robot that was mentioned in the video, for example. Although music is considered creative, 'good music' can actually be made into an algorithm. We've got a lot of psychological research about what patterns of sound people like and if you work within those patterns you can make a bot that will write music. That's not really creativity though. That bot will not be able to define a new music genre or field, rather it can create new sub-patterns in already existing fields.

    The long and short of current automation, and the near future of it, is that if there are 'patternable' tasks then it's likely we'll be able to create bots to perform them. These bots will be more reliable than humans but they most likely won't be as adaptable. That difference will be where the line is drawn for when it is desirable to use a human rather than a bot.

    Moving onto the economics side, I'll preface this with the fact that I know jack all about economics beyond my pay check and a balanced budget. My understanding, however, is that our economy is based on scarcity. Scarcity of a good, scarcity of labor, scarcity of resources. Robots, theoretically, can solve all three of these leaving us in a post-scarcity society where basic needs are taken care of. Two big problems, and one major obstacle occur to me.

    First, in a post-scarcity society what defines 'basic needs' and if you want more then how do you get more if there isn't a way of providing increased 'contribution' because all the work is being done by robots.

    Second, do we really want a society where all our needs are taken care of? The drive to 'survive' is usually listed as a huge factor in the drive to 'improve' in the psychology papers I've read and if we no longer have to worry about surviving how many of us will worry about 'improving?' I worry that in a society where we have 'everything we need' humanity will stagnate.

    Finally, the obstacle. Transitioning from a scarcity to a post-scarcity economy will be utterly chaotic and possibly impossible. We have an already established system where some have much more than others. Will they give that up willingly? Should they give that up? How do we determine how much everyone should have? What happens in the years, potentially decades, during which we have achieved post-scarcity in some areas and not others? At that point we'll still have a monetized economy and if increasing millions of people are going out of work with no new jobs on the horizon then we're going to be facing mass-starvation and almost certainly civil unrest because those millions aren't just going to waste away without a word.

    Ultimately I view a video like this one with skepticism because all it really does is make a lofty prediction and portray a grand idea without considering, or even covering, all the little details.

    Reply
    1. AoriiAorii

      But it's not meant to address the details. It's meant to open the scope for us to consider a possibility.
      The video does not answer how we should prepare for this, or when exactly it would come, or how quickly it would transition. The narrator is not trying to tell or predict the details of the future. Only to tell us that based on hindsight towards equivalent in history, it will inevitably arrive.
      The video has one and solely one purpose: to raise awareness. Because humans are good at solving problems, but only if we were aware of the problem in the first place.

      Reply
  16. Nyan Cat

    you know i go reddit for these kind of thing, there's a subreddit for debate you know

    Reply
  17. dan

    The video makes sense(in an extremly biased way), as long as the bot is used somewhere where there won't be a large problem when something goes wrong. The self driving cars, besides all the legal issues when an accident occurs, what about flat tires or unexpected mechanical faliures?

    Coffee shop, depends on if the customers are willing to buy, AND will depend on the price. I don't go to those, but I'm told that they're popular because of the atmosphere. Will the bot be able to contribute to the atmosphere? Or would it just be the same as one of those vending machines?

    It'll take time for the technology to develop and gain trust to be used. That should lead to a less painful transition compared to the video's implied "you're all out of a job tomorrow!"

    And the biggest problem with bots: The work isn't done when the power is out... "Sorry, can't sell you food because the power is out/cook bot broke." "But it's RIGHT THERE, and I have CASH!" "Nope, sorry. Power is out. No one here is trained to prepare the food."

    Reply
  18. Akugelap

    right now many ppl unemployable in my country and it not because of bot , it because their dont want to work at lower job than they want to work.many graduate still unemployed n blame industry or country authority for not helping them find job that they want. maybe some of u mad at me for saying this but it human nature to not look at himself for error but blame other thing for that error. for example in my country,restaurant n cleaner at public faculty mostly worker from Bangladesh that to come work in country n always get blame by unemployed ppl for stealing their job but when asked for them to replace bangladesh worker , they will say it not their job based their education level.so dont need to wait bot to replace human work to make many ppl dont have work, it already happen right now many ppl dont have job because be picky to choose job.even now i public servant , before that i do many misc. job n not all is legal job to support my parent.labor,cleaner ,catch freshwater fish to sell at local market n some more job before i got help from someone that have job in government when i working install electric wiring in his house.it just lucky that get my current job.

    Reply
    1. Reaper Phoenix

      Like me. I rather do freelance work than being cooped up in an office or do manual labor. My family is still on my case for not finding a steady mundane job.

      Reply
  19. Reaper Phoenix

    There are professions that are very unlikely to be replaced by machines, mostly the ones that requires empathy. Unless someone came up with an AI like Yui from SAO.

    Reply
  20. Nak

    Cool vid! It got me thinking. The way I see it, in the situation explained by the video, the ones taking everybody's "wages" aren't the robots, but the companies that implement more and more of them. It doesn't show the limit of comfort we can get from automation but the limit of capitalism and the evil that it is. The more somebody at the top earns, the more people below him are impoverished. It has always been true and this is just the limit of it we're seeing here. If everybody earns equally, does equal work and if nobody gains more than anybody else then you can see it as less and less work for everyone. In this case China will skate by and the USA is in serious trouble. I may be wrong, as I'm not a flawless robot(lol) but that's how I see it right now.

    I am soon to be a licensed architect(hoping), and I hope they don't invent a bot that can design buildings as well as we do. Architecture is a blend of art and science, and I see creativity as being immune to this problem(and in the video it only says that creative jobs are limited, not vulnerable). But I'm still worried by this.

    Reply
    1. Littel

      well , your idea of what economy should be is also quite flawed for a simple reason : first of all it's impossible for everyone to do the same job , some will always work better or harder than others. Then some people will not be intelligent enough or at least have the required talents for some jobs , so those will have to be done by the select ones that CAN do them. And finally , that's the problem the URSS encountered ( and which china slipped past , because exactly , their system is much much closer to capitalisme as what you seem to be thinking ) , it's that people who all earn equally whatever their job , won't have any motivation to actually do their job.

      Reply
      1. Nak

        Welp. Sorry I shouldn't have made political/economic examples as I have limited knowledge on it. It isn't the main point of my post though. I was merely pointing out what I saw as an inherent "kink" in humanity's logic.

        We hear the sentence "- Companies, in the future, will be forced to replace humans with robots", and we immediately blame the robots, or technology's fast pace, or we deny this and say "No, robots will never be as good as humans!". Don't you think it's ridiculous? In that sentence we just compared ourselves with a tool. We all know that tools are made to make people's lives easier and easier. They are not supposed to be made to replace us. The video talks about humans as if we're tools ourselves. It even goes so far as to state that the tools, which are made to make people's lives easier, will one day be better than humans. And we, the viewers, take all these in as if it's completely natural. What, then, is wrong with the picture? Why is it so natural for us now to think that we can be equated with a chunk of metal?

        It is overly idealistic, i know. And I have no right to preach. I just noticed it while watching this video after all.

        Reply
        1. KadiKadi Post author

          In the big picture named "economy", we are nothing but tools. Somewhat adaptable, but expensive tools. For better and worse.

          Reply
    2. AoriiAorii

      I would like to adjust one notion in your otherwise insightful argument:
      Capitalism isn't evil. It simply adheres to the laws of economic efficiency -- profitability.
      Would you describe the laws of physics as evil because it doesn't try to save men from dying?
      Same thing.
      It's not evil. It just doesn't care.
      And that's the problem.

      Reply
      1. Hmmm

        Plus capitalism takes advantage of our inner desirw which is to have more :P. If you are not satisfied which such economy based then you are feee to try to think of another way, but this is the best we got compared to command based and merchantilism economy

        Reply
  21. lol...

    lol... taking productivity out of the equation for now, the underlying point here, is that these machines, "exist" to make human life easier, and they are able to "exist" due to the fact that humans require consumption. The posted video, is being far too over-dramatic. Putting it simply, the most general concept, is that humans have unlimited needs and wants (relative scarcity), and they pay for these needs, and wants via, cash, this cash is earned from jobs. In saying so, by replacing a large amount of jobs, with robots with self intelligence, you would leave a bunch of people without pay, this leads to less consumption, as a result of less purchasing power, due to loss of disposable wage. What does this mean? Death (also over dramatic).

    now then for my own view. The video is completely idiotic. Nothing but doom-and-gloom rants, supported by foolish arguments. (WHY IS HE EVEN MENTIONING A DEPRESSION? ROBOTS TAKING OVER YOUR JOB DOES NOT EQUATE TO A DEPRESSION, WHY? BECAUSE IF THAT REALLY DID HAPPEN, THE OWNERS OF THE ROBOTS WOULD DIE THEMSELVES, AS THEIR POOL OF DAMN CUSTOMERS DWINDLES AND THE NEED FOR ROBOTS WOULD BECOME OBSOLETE)

    SIMPLY FREAKING PUT, ROBOTS TAKE YOUR JOB = NO MONEY TO SPEND = YOU STARVE = POPULATION REDUCTION = LESS NEED FOR PRODUCTION = ROBOTS ALSO BECOME OBSOLETE = LOL ARMAGEDDON (and to those that argue otherwise, answer me this, if a bunch of people lose ther jobs, then sales will go down, and production with it, thus the available pool of customers will continue to dwindle, which business-man in their right mind would do this? there's an extremely large difference between desire for productivity, and desire for self ruin.)

    Reply
    1. lol...

      btw, krytyk... any tips on japanese? i know about things like common radicals, e.g. 言 on the left (or sometimes bottom)=words, talking, writing, etc. are but how can i do it faster? having to resort to jisho and using radicals gets to be annoying... or is it just practice practice practice? i mean i've been learning for half a year, and it takes me half an hour to translate 1 page of an ln....

      or around 10 mins for a small 4-koma...

      QQ

      Reply
    2. Nak

      Calm down dude. I think everybody is aware of how supply and demand works so no need to go all caps on us. :)

      Reply
    3. AoriiAorii

      Supply and demand isn't that simple though, nor that elastic. I had the fortune of going into "intermediate economics" back when I wanted an econ minor. The first thing they told us when we entered is: all that stuff you learned in intro / basic economics? they're all lies because they presume you work in a microcosm. That's not how the world functions.
      For example, short-run supply demand is very different from long-run supply demand -- which has very little elasticity.

      The robot revolution will reduce the need for humans and therefore production, but it doesn't come fast enough. It will take decades for humanity's population to lower, generations if you put cultural conservatism into perspective. What do we do with all those people during those decades? send them to battlefields or drop nukes on them? The need to "cull the human population" to restore balance via invisible hand is a chilling thought.

      Reply
  22. Hmmm

    I think of the future as something akin to mahouka without the magic. A public personnal transportation will be created but private transportation will still used. Of course a lot job will be is lost but new jobs will also be created. made. Also if technology manage to advance that far then it is pretty safe to assume the vr like SAO would be possible. Moreover the video points out that it machine would be able to create better stuff and that opens the door to real space exploration and colonization. Seriously, there is no real threat to humans but just a key to infinite possibilities. I for one is looking forward for the next decade or so since i watched the video.

    Reply
  23. Paul Schuster

    I think that the video is mostly overblown. Customer service is still the most important part of any front facing industry out there.

    The video has in it an example of burista's having there job taken away by a machine, I work in a similar industry being a bartender. There is almost no reason why my job would be taken over by a box that takes orders (and yes they do exist) for a couple of reasons. First the job is not easy and making an honest drink is the easiest part of the job. Second most bars/coffee shops run with as few people as possible, as a matter of fact during the day I may be the only employee there. The issue with this is even with the drink-o-bot 2k there I would still have to be there to handle everything that makes a brick and mortar store run. Lastly just from this example it would not replace a reasonable bartender as with out a person there to talk to, it would be nothing different than getting a can of pop from a vending machine. People go to bars and coffee shops to have communal experiences that allow them to feel connected to other people, even if they are not talking to anyone around them. With out a bartender or a burista it is what is shown in the video, a very fancy vending machine, the same type that has not put bars or coffee shops out of business when they existed in the past.

    Some jobs will of course be lost. In the back of the house there may be no need for line cooks, or dishwashers. Though to be honest that has been mostly true for a while now. A donut shop might not have had anyone doing a bake in a decade, there is no one prepping the greens for a salad as they come in a bag, and the house dressing says sysco on the side of the bag.

    Reply
    1. AoriiAorii

      For now, I agree. Humans prefer to see another human face. That's part of psychology.
      But psychology also evolves faster than biology. People will get used to robots a lot faster than most people believe, just as how quickly we got used to the internet. Cheap coffee shops will attract customers, especially in today's flailing economy, and people will find alternate mediums of social interaction. We adapt to see our needs, after all.
      Human service may eventually become a "high class commodity" just like horses. Unfortunately, high class doesn't have a high demand. Some will benefit. Most will lose their jobs.
      The best point the video makes on that is production has never been about the "best". The first law of engineering is cost vs quality. We could make a product that never fails and never kill anyone through failure. But would that be worth it? Not really. (Yes, I was actually taught this in Engineering school. Is this ethical? that's up to debate. But it's Fact) Quantitative efficiency is what most people care about. Quality is a want, not a need.

      Reply
  24. Owl

    Well, think the general consensus here is that you got conned Kadi. It's not a problem, just someone being anti-computer or anti-machine.

    Reply
    1. Shiku

      Hmmmm... I think "conned" isn't the right word.

      If we think of "con"... it means a huge loss and no merits/previlege/advantage but that did not occur.
      Kadi-san or us have no advantage nor disadvantage about the topic to begin with. It's more like "joining a different group", pros, neutral, and the cons' side gained a valuable info in the topic but in a different perspective.

      Reply
      1. Owl

        I don't think being conned is dependent on material gain or loss, just the fact that someone was led to believe something untrue. I know the common usage involves scams, but that is due to the prevalence of scams, not the usage of the word.

        Reply
        1. Littel

          in that case he hasnt gotten conned either , since what he was led to believe wasn't really false , but was only a small part of the truth :) And let's not forget everything the video or our comments said are prediction , everyone has his own idea but no one's sure about what will happen next. Who knows if in a decade there won't be a terrible outbreak of whaterveryouwantsit and that kills off a third of the population , clearing the concerns for unemployment during quite some time ? ^^

          Reply
  25. Shiku

    After watching the video... To be honest with you, The topic is a one-sided POV about THE advantage of machines, well most part of it is mostly true... But the point is, the explanations are only half-truth or some may call it "concealed", The topic merely explaining the HUGE ADVANTAGE of machines compared to humans by pointing various mistakes or disadvantages in area.

    I may be just a computer programmer student, but I feel the video is utterly unjustified beatdown.
    The video didn't go to the pros and cons of the topic and obviously merciless beating humans' POV.

    And oh, In the Self Learning of bots topic... Achieving general intellegence that humans have, that will be just like a star of admiration, bots maybe can self learn itself, train itself, and corrects itself. That will be just a replica image of what we call "perfection". As of now, none really explain the limits of human mind, it will be stay just a fiction and a theory.

    Its not I disagree completely about the video... its just like that the video is talking "Machines are supreme than human" feeling. If the video is talking the pros and cons one by one, I may believe it... but it's not.

    But thanks though ^^

    Reply
    1. AoriiAorii

      You're right. It's very one-sided. But "awareness" videos are usually made to be alarmist.
      Pegasystems already has a product that can cut most software engineering teams from hundreds down to just a dozen or so through the use of preprogrammed and self-programming modules. That's scary enough for a starter of what's to come.
      They don't need to replace all of us to create chaos. They just need to replace enough. Machines may not be perfect, but are they as flawed as most humans? Those flaws are shrinking faster than our brains improve, that's for sure.

      Reply
  26. Ekmo

    So if I only want to sell 10 cars a day, I could hire that 1 person, or 50 Baxters for half the cost of that 1 person.

    I would be really interested to see how technology will cope with language interaction. I work in the healthcare field myself, and there are so many ways the word "Fine" can be interpreted. Some don't even involve listening to that word.

    The problem with humans is irrationality. Until computers can understand that, total replacement is pretty far off.

    Reply
    1. Owl

      Don't forget the 6 maintenance staff you need to support the "Baxters". And the IT support. And since the IT and support staff are specialists, their cost is going to be pretty high too!

      "Hi, my Baxter blue screened, and I can't find the Ctrl-Alt-Del button."
      "Woo... sounds bad, that's going to cost you..." :)

      I'm also a bit into healthcare, and since it is a small lab, I need to do some invoicing as well. Cost for a technician for repairs is 800 bucks for a job that took him 8 hours. 100 dollars an hour. Sometimes, when I see that, I think I'm in the wrong job. I only get 10 an hour!

      Reply
      1. Calvin

        I was in IT as a contractor. That technician only got $10 to $20 of that $100/hour. The rest went to his company and taxes.

        Reply
    2. Littel

      language algorythms actually work fairly well already ( as i said .... FAIRLY ^^ ) , the problem is more about the complexity level of the "thinking" it can achieve ^^ . But there's no way , at least no way anyone could think of at this time , to make a machine integrate the principle of irrationality :/ at most you could make it ignore any irrational behavior ^^"

      Reply
  27. Owl

    It's an alarmist video. Worst case, people become programmers instead of field workers. All this really is old. Even the term "Robot" was from a Russian story about mechanical labour taking over human beings. This is during the 1900s I think.

    I'm with viracom. If "robots" and "computers" are all so smart, why is Google Translate so terrible! :)

    "Meet Baxter, he's going to take over your job!"
    "How many cars can he make a day?"
    "0.2"
    "So he takes 5 days to build a car?"
    "Yes"
    "We do 20 a day!"
    "But he's 100x cheaper!"
    "Only if you want to sell a car every 5 days!"

    Reply
      1. Owl

        Can you imagine the drop in food production if people went wholesale with that idea? Sometimes, immediate output is critical, especially perishable foodstuff. You'll need huge amounts of machines to do the job of just one person in a timely manner.

        Hell, PCs are pretty common now, but you haven't seen anyone fired and replaced by a PC yet have you? Worst case "Looking for people who can use MS Word". "Looking for IT personnel who can troubleshoot IS" "Looking for Linix Programmer" etc. It opens more and different jobs rather than lock out people, which is also a good thing.

        Think "Irish Potato Famine". You focus your industry too much in one area, if anything happens, you're screwed. At least with diversity, one sector can cover for another for a short while.

        Reply
    1. Spade

      You put that in math, baxter is 100x cheaper and he make at least 5 per day. So if you have 100 baxter which would be equivalent to a normal cost of labor then he would be able to create 500 per day compared that to measely .

      Reply
    2. Littel

      first : it wouldnt make sense to take the robot if overall the price would be equivalent ( 100x0.2=20 ) but the speed still a hundred times inferior ^^
      second : sadly , the car industry is already nearly fully automatised ; so there's no use for baxter here , and even if he got used here he wouldnt be doing a whole car on his all ( because he can't ) , he'd just do a specific task , which would be about the same thing as what workers did before and automatized chains of production do today

      Reply
  28. Reaper Phoenix

    Economy is also about supply and demand. If people are jobless and broke, there will be less demand for products and services. Also with nothing to do, people are more likely to resort to crime due to boredom.
    I hope the people who are developing these technologies consider the socio economic impact of their work.

    Reply
    1. AoriiAorii

      Just as much as the developers of the nuclear bomb considered it's implications fully.
      "Now I am become death, the destroyer of worlds." -- Oppenheimer after he exploded the first bomb.
      Engineers aren't as ethical or farsighted as people would like them to be. And businessmen far worse xD

      Reply
  29. Hikari_

    anyway, I Like the catchphrase: "Justise and Fluffinesse, Darknesse and Dawn" you mentioned there...

    Reply
  30. maqui

    I didnt viewed this video but read the comments and saw the idea behind it and i dont agree , the idea of mass automatization isnt really all that beneficial to the owners of said robots , because if all labor jobs are going to be taken how the money flows, no employees no salary no money so the products vreated by said robots will not bring money so said complete automatization isnt so benefictional.You con think this as the same ways as labor laws in your country.
    Ps. Sorry for poor english, it isnt my first language.

    Reply
  31. Nissarin

    I've to agree with Littel on this one.. it will take a lot of time before we will be able to put our trust in robots. In my opinion it's impossible in foreseeable future, because in order for them to deal with "unexpected" we would need to construct a "real" AI - thinking machine.
    We still have no idea how to do it properly and (most importantly) we lack proper hardware to do so, it's harder and harder to make any faster CPU, quantum computing won't leave the labs for decades (and only selected algorithms can take advantage of it).
    Will it happen someday ? Maybe. Hopefully at that time we will have money free, utopian society... but having "thinking" machines around is scary in it's own way, not to mention leads to some moral dilemmas (if they are thinking for themselves they should have equal rights) .
    For now list of things which can be automated with todays (even tomorrows) technology and make sense (economically) is pretty short... when automated vacuum cleaners learn to avoid dog's we might be in trouble though.

    Reply
    1. AoriiAorii

      It also took a lot of time before we trust internet sources, or purely digital storage, or any other medium of technology.
      The time usually ends up being only a decade.
      Psychology also evolves, and it evolves faster than biology too.
      Besides, trust between humans do not come easily either, mostly because of our inherent flaw in being illogical and emotional creatures.

      Reply
  32. babbo3d

    it looks interesting thought i would like to see that future (not the unemployment one, but autonomous bots that are general purpose).

    Reply
  33. Sanngrior

    hmm, sorry everyone. my future job kinda involves making that situation occur sooner than it would otherwise :s

    i just tend to trust a well made machine over the occasional human stupidity :(

    atleast for now they arent smart enough

    Reply
  34. nierror

    I don't see this as a problem at all if machines do everyone's jobs then we'll finally get to live in a sci-fi world where everyone is on permanent vacation. Isn't a post-scarcity society what everyone has wanted since the dawn of time?

    Reply
    1. Bagelson

      Yeah, automation in itself isn't the problem, and I don't interpret this as the point of the video. The problem is how to deal with it.

      When automation started becoming more common a hundred years ago the idea was that with the help of machines, people could work less; when the job can be done in a quarter of the time, people only have to work a quarter of the time and then go home. Of course, this did not happen. Instead the companies fired three quarters of their workers and could maintain the same productivity at a fraction of the personnel costs.

      Automation is inevitable, so there's no point in seeing it as a problem. But we as a society has to be ready to embrace that, which at present we are not.

      Reply
  35. Viracom

    Still waiting for that automatic translation bots that are actually readable, so i guess there will be at least people whining for LN translations for a while.
    And seriously if there is something that humans are good at it's - breaking stuff - so i doubt humans will be out of jobs anytime soon since most robots/ autmatic software isn't really that great in protecting itself - from us.

    Reply
  36. Littel

    ho yeah , forgot an important point in my previous comment , and that is we're reaching the pysical limits of miniturizing computers already ^^ so the neighborhood thing is not something only true right now , it will stay true as of now ^^"

    Reply
      1. Littel

        yah those are kind of the exception , but honestly , the people able to program those can be counted on your fingers , and this number will barely go up considering the complexity. moreover , that's only considering the act of programming them. actually using it to make complex algorythms like AIs seem quite fantasy-like currently ^^

        Reply
  37. fishsandwich

    I did watch the video and it is alarmist to say the least. While automation, robots, and computers are important tools that are highly efficient, They lack human intelligence in the problem solving department. As another commenter noted, robots are purely reactive and scripted. It cannot predict next steps, even if it can learn, it can only learn through trial and error. Would you let the robot experiment with your prescription or how about how many bolts it takes to secure your car's steering column? Even if they can learn, they are nowhere near adequate yet. Even by the best estimates, we are generations away. Right now for the foreseeable future these tools will most remain just that, tool to enhance human efficiency.

    Reply
    1. KadiKadi Post author

      But even if it takes 30 years to come... At least I'll still be a working member of society then. I hope.

      Reply
    2. AoriiAorii

      Guess what? Most jobs are also purely reactive and scripted.
      Being able to adapt and decide on the fly is a talent sought in humans, not a talent every human has. Almost every decision-making role always falls back on experience, because people are not persuaded by intuition, they are persuaded by hard data. In companies you are expected to compare and contrast to the market, to past accounting history, to simulated projections, etc. Good decision-making is a very logical, very procedural method of analyzing risk vs gains, and anything purely logical is automatable.
      Humans rely on trial and error to learn just as much. The only problem? You can't copy and paste a human's mind (yet; it's also an ethical issue). Teaching just isn't the same, and certainly not as efficient.
      You can do so with a machine.

      Reply
  38. Calvin

    It's more a problem for First World nations, where labour costs are high. Even then, it's not a new phenomena. After all, there's already outsourcing to 3rd World nations, immigrant labour, computerization, mechanization, etc. Furthermore, there is an assumption of continued cheap energy and cheap materials (plastics, which rely on cheap oil, and especially the rare earth stuff needed for more advanced micro-electronic and computer circuits.) This assumption is more obviously becoming a fallacy as time goes on.

    Frankly, I'm more worried about oil, energy and water shortages, not to mention global warming. The state of the world economy, as well as the modern economic doctrine which posits infinite growth like a perpetual motion machine and decouples productivity from limited availability of raw materials is another area of concern. In other words, most of the world outside of the US and a few European countries have bigger worries than losing jobs to robots.

    Reply
  39. Littel

    hey Kadi , I hadn't seen this video yet , through those are topics i hear about a lot since I'm currently in an engineering school , and specializing in automatisation and robotics. So i though i could pinch my grain of salt in there , but of course it's the same as the video , no one's force to believe anything i say ^^ let me add as well that I'm not an english native speaker , so please excuse this pitiful me for any mistakes or unexistent words used .
    Well for starters let's be honest : yes , there are a lot of jobs , way more than anyone could think that can currently or in a near future be made by bots. auto learning bots are also a reality, I've read enough papers about them and their uses to confirm that ^^. fortunately for some of you , they won't ever be able to do everything by themselves.
    The main point here , and the big fault I can find in this video , is that what will hinder greatly the arrival and developement of those will be * rolling drums * ........ humans . The video took it as a purely economical efficiency point of view, and in that regard there is no telling how boudlessly better bots are compared to humans, at least how their potential is, but the fact that " they only need to not be as bad as humans to take their place" is quite a big misconception. The only form of automatisation that can completely run without human help , except of course for programmers and maintenance , are the ones used in industrial mass producing chains , but those took quite a while to take the place of humans for a simple reason : they had to be flawless before being usable.
    Let's take the exemple of auto cars. Those things are absolutely marvelous. Just take into account the fact that over 95% of trafic jams are caused by human behavior , and that if the trajectory of every vehicule was centralized and coordinated , most of them could be resolved by simply reducing the speed of all cars by 10% or a better positionning on the road. BUT , there's a huge problem , and that only comes when the system does not fulfill it's intended purpose. That is : who will you blame ? humans have no trouble blaming other humans , but how will you blame a machine ? how will you have your " revenge " ? of course when talking about trafic jams it seems silly , but when talking about an auto car's optic detection system not working well because , let's say a leaf covered it , and it runs over a human , what will happen then ? will people satisfy themselves with " ho well , that happenned but if you see the picture as a whole , there's a lot less deaths then there used to be " ? the whole point of remplacing people by machines is putting your trust in them . machines have a disadvantage compared to human ; that is they are not able to react in a time of crisis. intelligent bots are here , but those are only specialized. the brain capacity of a human is so much larger compared to what a computer can achieve that a bot could never compare , except if he was the size of a freaking neighborhood , and that would be assuming we had enough skilled programmers to make it work once the material infrastructure needed is there ^^ a bot can only react to " expected " situations , which means either situations programmed in it by humans or situations from which it has learned from , but there always happens to be " unexpected" ones , and let me assure you that in this case , all and every bot will go completely crazy , taking non sensical actions, and those actions would make people not able to trust bots , meaning boycotting them ^^ of course other jobs will disapear , replaced by bots , but there's a limit to it , and this limit is much lower than that video would let you think

    Sorry for the long post , sorry as well if this was somewhat unclear and messy since i was too lazy to draft a text and rework it until it was perfect ^^ anyway if you reached here then i praise your tenacity in reading this aweful text :D hoping this has calmed down some of your worries

    Reply
    1. AoriiAorii

      Actually, warmgaming AI and algorithms has already shown: machines handle stressful, critical situation better than most men. When disaster strikes, they will fall back on their experience just like most veteran commanders do, except machines can come with that experience where it takes years and possibly many failures for a human to learn the same.
      It's true that machines can't reach as high as man (without true sentience, anyway). But the point of the video is that they don't need to replace ALL of us. Just replacing enough of us (45%) will easily destabilize society.
      45% is easy.

      Also, as an engineer who worked in company for several years and is taking managerial classes: if you think things are not completely and totally driven by economies of efficiency and profitability, prepare for a rude awakening. How market work is very different from how we would like it to work -- so much that it's kind of disgusting.

      Reply
      1. Littel

        I'm not denying the market does not work on efficiency and profitabillity , but it's a fact that comapnies and products can completely disapear from the market if their products lose the trust of their users. Also , the game AIs are quite a good exemple of what I was saying earlier : they are awefully specialized ^^ since the programmers know what can happen in the game , they make it so they can react to those. ever seen the reaction of an AI to someone using cheats ? well either they ignore it and act as if there was none , or go crazy ^^ and now just imagine an AI in a normal racing game , and then suddenly the road before the car explodes. any idea how it would react ? that's what i meant by not being able to handle crisis :)

        Reply
      2. Littel

        And overall , the main problem I pointed out was the blame ^^ if you could shift the blame of those accidents on the maker of those machines , they wouldn't produce them anymore. But if you can't put the blame on anyone , then I'm pretty sure the blame will just fall on the machines in general and they won't be used anymore. efficiency aside , profitability would soon become pretty bad

        Reply
  40. Loco15

    Hah, I live in Venezuela, the thought of machines doing that here before I die is almost as unthinkable as the idea of playing volleyball in space while wearing swimsuits.

    Reply