Jump to content

Welcome to FutureTimeline.forum
Register now to gain access to all of our features. Once registered and logged in, you will be able to create topics, post replies to existing threads, give reputation to your fellow members, get your own private messenger, post status updates, manage your profile and so much more. If you already have an account, login here - otherwise create an account for free today!

These ads will disappear if you register on the forum

Photo

Library of Futurology Discussions

Futurology philosophy Singularity artificial intelligence energy automation politics war society economy

  • Please log in to reply
2 replies to this topic

#1
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,148 posts
  • LocationAnur Margidda

Here is a thread meant to act as a hub for all those wonderful long, thought-out, thought-provoking futurology posts from Reddit, Tumblr, 4chan, other forums, or even this forum if you will. No official opinion pieces/ articles— those go in their respective News and Discussions threads or as their own topic on another subforum. Basically, they have to be comments or posts.

Always link to the discussion as well— that's the whole point, so we can also see the accompanying discussion.
 
Inspired by CthulhusCallerID's two posts, as I'll show below. Use these as a guide for what to post.
 
"Will Robots Create New Jobs When They Take Over Existing Ones?"

I'd be a little curious for both of you to unpack your reasoning.
For my part, the writer mentions agricultural jobs. The tractor revolutionized farming, this is true, but what the tractor really replaced was the horse and the mule. People moved from those jobs into factory jobs (low skill manual farm labor, to low skill manual industrial labor). Mechanical power replaced muscle power, but it did not replace fine motor skills or the human brain.
When we started losing factory work to outsourcing and automation, people moved from factory jobs to service jobs (another low skills to low skills transition). Some fine motor skill tasks were replaced, but not by a machine that was flexible to perform a variety of tasks, so there were high up front costs for narrow goals. These machines could not interact in any social settings (no natural language processing) so their were a variety of tasks that require mental flexibility that are hard to cheat.
Now we're on the cusp of creating machines with our fine motor skills, similar visual capabilities, that can be trained to perform a variety of tasks, and, while still fragile, the ability to parse information and give responses to natural language questions (this last is still a little ways down the road, but you can see the beginnings in Watson, some translation software, and in the scaled down devices like siri).
I suspect jobs will be dramatically impacted. First, the workers who are going to lose their jobs first are the ones who will have the most difficult time retraining to work in a more high skills market place. People who lack the income to go back to college and who may not thrive in that kind of environment to begin with. Not every truck driver is likely to smoothly transition into being a programmer, not every retail worker is likely to transition into mechanical engineering. These job changes are much more significant than the transition of farm hands to factory workers.
Secondly, when mechanical power replaced muscle power, we didn't find other uses for horses. There was something on the order of 1.5 million working horses around London prior to the industrial revolution. There are now a thousand or so. When a reasonably good AI (even if it's narrow) can do the work of a person at a price point that is less than the cost of employing a person, it will replace that person. The cost of training that AI falls, as once the first one is serviceable at a task, they are all serviceable at that task. Once one is great, they will all be great. Compare this to humans, the transmission of skills and ideas through education is comparatively clumsy with uneven results.
The third point I'd like to bring up is that there are practical limits to the benefits (as far as employment is concerned) of productivity gains. The industrial revolution also saw certain goods drop in prices which expanded the markets of who could buy those goods. At the same time, colonialism (which I'm not saying was a good thing) expanded the reach of trade. Goods made in England could be sold in India and vice versa. So the market for where goods could be sold expanded dramatically, even faster than the increased production could accommodate. Our supply chains have reached (virtually) every corner of the globe at this point. There aren't vast continents full of people waiting to be discovered. Now our markets can only be expanded by 1) helping the impoverished develop economically- historically we haven't been great at doing this. Or, 2) population increases. If we're increasing our productivity faster than people are rising out of poverty and faster than the population is expanding, then we can expect to see efficiency gains be offset by job loses which would reduce poverty reduction.
Additionally, there are other disruptive technologies that are going to make it more difficult to see a future where employment is a viable options for most people. 3D printing is going to hit manufacturing very hard. Virtual and augmented reality may hit a variety of business including airlines and hotels (why travel for 10 hours each way when you can pop on your vr suit and feel as though you're in Hawaii right now? Why send your employees to the big meeting in Fresno when everyone can pop on their google glasses and it's like they're all in the same conference room?).

 
"Do you worry about collapsing economies during the times of rapid change?"

I do, yes. There's a lag between when the cost to produce a product falls and when the prices fall (economists might say that prices are sticky). There's also a lag between when people are laid off because of outsourcing or technological unemployment and when the economy creates new jobs or moves labor to where it's needed most. Since technological innovation is accelerating while the time it takes humans to retrain for news jobs remains relatively fixed, I think we can predict that more and more people will experience unemployment, with more people experiencing it, there's greater competition for the jobs that open up, this decreases anyone's likelihood of securing a new job and increase the length they're with out a job (on average), this will have an overall wage suppression effect and create a feedback cycle throughout the system. Fewer people will be buying things because they lack wages, this creates a greater incentive to cut costs while maintaining a high profit margin, which means cutting jobs (lowering prices is usually the last option, as you can maintain your margin while selling less product by keeping prices steady but reducing the costs of production).
Eventually, AI, fast approaching human level intelligence, begins to truly decimate the remaining job market. There is nothing, potentially including creative tasks, that we can do that can't be emulated or simulated. The advantages of an AI are huge compared to an unaugmented human. For starters, once one is programed, trained or learns to do a task, all of them can then do that task, meaning as new kinds of work open up humans can only fill those roles until an AI has mastered them, at which point each additional human that needs to be trained would better be replaced by an AI which is immediately competent and requires no further training. Additionally as energy costs fall, the upkeep of the machines becomes smaller and smaller as do the costs of the machines themselves.
In the meantime these factors create more pressure on the human labor-force to differentiate themselves from one another, this can best be down through education which means increased debt, and less spending on other things (cars, housing, entertainment, etc), which, of course accelerates the feedback cycle.
Additionally there are companies within the economies which are best positioned to succeed in the changing world may be able to cut costs or otherwise increase efficiency to such a degree that they can dramatically out compete foreign competitors even within their own markets, this in turn could either trigger protectionist policies (almost always bad) or create fewer niches in the market for people inside those countries to use their labor pool (not inherently bad if they can find other, better avenues for using this labor, but it could extend periods of unemployment and that's bad). This concept is outlined for other reasons in Baseball Between the Numbers on the chapter of why cities subsidizing sports stadiums is almost always a bad idea.
As a final point, the first wave of people whose jobs will completely cease to exit will be drivers (truck drivers, taxi drivers, potentially pilots and other large machinery) and we should start seeing that in the next ten years or so. They're an interesting case for a few reasons. The first is that they will be replaced almost whole sale, as opposed to seeing efficiency gains incrementally making us need fewer and fewer drivers, meaning one day there are virtually no self driving cars and the next each new car is self driving. This is much, much more sudden shift than what happened during the industrial revolution which took place over the course of 60 years, giving people generations to move into different sectors of the economy. Two, because it's whole sale and sudden, this will be much more visible and less esoteric than some software which reduces the amount of effort it takes a white collar worker to preform some task or some robot which frees up wielders to address the areas that take more skill. Its visibility will drive a public conversation we haven't had before. Finally, because this is a group of people who are generally economically disadvantaged (or, at the very least, not privileged) and generally aren't well educated, retraining them is going to be difficult, and they're unlikely to find jobs that pay as well if they are able to find jobs at all.
The truckers are going to be the new Luddites. The debate we have and the policies we develop to deal with their unemployment is probably going to set the tone for how we are treated as accelerating change devalues our contributions as well. What I hope will happen is that we think outside the box about how best to help us transition into a post scarcity economy. What I fear will happen will be powerful interests run roughshod over everyone and impose laws which impose false scarcity on the rest of us (think of the more draconian copyright law enforcement and artificially high prices of digital media). Perhaps they're given enough to survive but not enough to retrain themselves or get access to the latest medical technologies that could augment their intelligence and therefore their competitiveness, rendering them a permanent underclass.
Economic instability frequently does led to social upheaval. Currently Greece's NeoNazi party is polling at 11.5% and support is even higher amongst non-voting teens (grain of salt, read about this a while ago in the economist), suggesting that, unlike the US, in Greece the extreme right wing is attractive to the youth, who are unemployed in record numbers (50+% and growing fast). In Spain, Franco's legacy is too strong for the far right to catch on, but the 15-M Movement shows how deeply dissatisfied they are. If things continue to be bad there, they could gravitate towards some other extreme, such as communism, in an effort to alleviate their suffering.
TL;DR: Change is accelerating but our abilities to retrain or reeducate ourselves are not (at least for unaugmented humans) and come at a cost. The change is going to happen much faster than the 60 year industrial revolution. AI will probably make most people's intellectual labor progressively more obsolete as we approach the singularity. Prices are sticky and will remain as high as possible for as long as possible. Yes, I'm worried though not of a total collapse.


So basically anything that's a bit long (at least more than a paragraph) and really makes you think, has many statistics, and is at least legible.


  • Casey, sasuke2490 and FrogCAT like this
Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!

#2
sasuke2490

sasuke2490

    veteran gamer

  • Members
  • PipPipPipPipPip
  • 464 posts

The human genome project began in 1993. In 1997 only one percent of the human genome had been sequenced. The experts in computers and processing, and those in human genetics said at that rate it would take roughly 700 years to sequence the entire human genome. Based on the current technology of that year they were correct. Yet the human genome was completely sequenced by 2005. Twelve years. Today we sequence more than 10,000 human genomes a year. (oh also. The first human genome sequenced cost about 3.5 billion dollars. Today it costs less than 500 dollars each and dropping.) The experts did not understand the concept of exponential increases in data processing power.

I think the same thing holds true here as well. Almost daily our medicine become more and more information technology based. I bet we come up with some pretty doggone amazing things within the next 10 years that we could not anticipate today 2016.

(That first paragraph is a bit of a paraphrase of Raymond Kurzweil's "The Singularity Is Near")

 

How do i get rid of this whtie stuff.


  • Casey likes this

https://www.instagc.com/scorch9722

Use my link for free steam wallet codes if you want


#3
Yuli Ban

Yuli Ban

    Nadsat Brat

  • Moderators
  • PipPipPipPipPipPipPipPipPipPipPip
  • 17,148 posts
  • LocationAnur Margidda

How do i get rid of this whtie stuff.

FGhAp53.jpg

Results!



The human genome project began in 1993. In 1997 only one percent of the human genome had been sequenced. The experts in computers and processing, and those in human genetics said at that rate it would take roughly 700 years to sequence the entire human genome. Based on the current technology of that year they were correct. Yet the human genome was completely sequenced by 2005. Twelve years. Today we sequence more than 10,000 human genomes a year. (oh also. The first human genome sequenced cost about 3.5 billion dollars. Today it costs less than 500 dollars each and dropping.) The experts did not understand the concept of exponential increases in data processing power.
I think the same thing holds true here as well. Almost daily our medicine become more and more information technology based. I bet we come up with some pretty doggone amazing things within the next 10 years that we could not anticipate today 2016.
(That first paragraph is a bit of a paraphrase of Raymond Kurzweil's "The Singularity Is Near")


Of course, it removes all formatting. So if there are any italics, bolds, or underlines that you want to keep, you have to manually put it back in.


  • caltrek and BasilBerylium like this
Nobody's gonna take my drone, I'm gonna fly miles far too high!
Nobody gonna beat my drone, it's gonna shoot into the sky!





Also tagged with one or more of these keywords: Futurology, philosophy, Singularity, artificial intelligence, energy, automation, politics, war, society, economy

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users