Automation And Jobs- Why It Really Is Different This Time.

[spyguy]

Theres a piece of a farm robot in the FT:

https://www.ft.com/content/beed97d2-28ff-11e7-bc4b-5528796fe35c#comments

Like the smart idea to import loads of Mipuiri labour in the 70s saw us loose our textile production and gain ~2m benefit sucking Muslims, the idea of importing loads of EE to harvest our crops has put UK farming 20 years behind.

The comments are a hoot.

' Will retired Lincolnshire folk be capable or interested or willing to operate and manage these machines.  Skeptical.  I think engineers and technicians from eastern Europe will be essential. '

A Romanian agri tech is some one who brushes down a horse FFS.

 

 

[scepticus]

The Myth of a Superhuman AI

https://backchannel.com/the-myth-of-a-superhuman-ai-59282b686c62

[Sawitcoming]

There is another very important aspect to this that makes it different this time and that is the pace of change. The pace of the first industrial revolution was slower and even though it was disruptive there was more time for people and importantly society to adjust. There is going to be more change in the next 20 years than there was in the last 400 years. 

[Errol]

The main difference is that eventually we (humans) will have nowhere to go. We will be the equivalent of the horse after the industrial revolution.

 

[Riedquat]

1 hour ago, Sawitcoming said:

There is another very important aspect to this that makes it different this time and that is the pace of change. The pace of the first industrial revolution was slower and even though it was disruptive there was more time for people and importantly society to adjust. There is going to be more change in the next 20 years than there was in the last 400 years. 

400 years ago the world was a completely different place. 20 years from now it'll be different from now but not to the same utterly unrecognisable extent (assuming you don't mean WWIII is going to kick off and we'll all be nuked to oblivion, the way things are looking now you never know)

[Sawitcoming]

25 minutes ago, Riedquat said:

400 years ago the world was a completely different place. 20 years from now it'll be different from now but not to the same utterly unrecognisable extent (assuming you don't mean WWIII is going to kick off and we'll all be nuked to oblivion, the way things are looking now you never know)

No I don't mean that. Putting that aside I mean that technology grows exponentially (moors law etc) and we are reaching an apex in that curve where the exonential growth in technology is accelerating at a significant rate. 

So far the future and its sci do technologies have been  a bit well... meh. However the next few years will begin to feel more dramatically sci-fi.

General ai such as deepmind  referred to in this post is probably one of the most significant   

But it is not just the individual technologies. It is the combination of emerging technologies that is significant  

 

 

 

Edited April 26, 2017 by Sawitcoming

[Riedquat]

An awful lot of those technological changes are really gimmicks though, so it remains to be seen what their long-range impact will be (probably pointlessly self-destructive for us). The big changes have largely already happened (in the first world at least) - the real problems that affect being able to live, without much real worry about the basics, have been solved for a long time now, and that's the real difference between now and 400 years ago, not wether or not you've got a robot to wipe your backside and tie your shoelaces for you (and no doubt in the nearish future people will be saying how awful life must've been without that robot).

Edited April 26, 2017 by Riedquat

[Brendan110_0]

27 minutes ago, Riedquat said:

An awful lot of those technological changes are really gimmicks though, so it remains to be seen what their long-range impact will be (probably pointlessly self-destructive for us). The big changes have largely already happened (in the first world at least) - the real problems that affect being able to live, without much real worry about the basics, have been solved for a long time now, and that's the real difference between now and 400 years ago, not wether or not you've got a robot to wipe your backside and tie your shoelaces for you (and no doubt in the nearish future people will be saying how awful life must've been without that robot).

Japanese toilet seat will sort that out (not shoe laces), maybe it'll also diagnose medical problems with an AI nose built in and e-mail your doctor for a check up.

[Sawitcoming]

4 hours ago, Errol said:

The main difference is that eventually we (humans) will have nowhere to go. We will be the equivalent of the horse after the industrial revolution.

 

So earlier in the year I was watching a lot of videos from the previous two years of Davos World Economic Forum and a big topic of conversation was Digital Transformation (Improvement and automation of services through digital technology). All of the CEOs there realise that they can no longer sit on the sidelines thinking its ok we are too big. The music industry went down that way and the other industries know its going to happen to them so they are moving on it and speed of transformation is quoted as being one of the big factors for success.

They all recognise that a big part of this automation will replace many jobs and they all recognise the issue. However, it is clear that nobody is acting on how to manage the change. UBI is talked about, but while it is recognised we will need entirely different social structures to deal with the fact that there will be large numbers of redundant people, it is clear nobody is doing anything about it. Automation is in itself not a bad thing. It is what our masters think our (ordinary people) value / role is now that we are no longer required resources.

 

 

[rollover]

 

Quote

 

Hyperrealistic robot

Lumidolls opened their first European brothel in Barcelona recently and now are looking to open more across the continent. A spokesperson said: “We are currently in the process of expanding and looking for more franchisees in other countries. The firm’s website says: “These are totally realistic dolls, both in the movement of their joints to the touch of their skin that will allow you to fulfil all your fantasies.” It adds: “These sex dolls will make the experience more pleasurable, exciting and erotic.” The spokesperson added: “With Lumidolls, customers can perform any of their sexual fantasies they do not dare to do with a woman. Express

 

I think prices will collapse soon for this kind of services.

[rollover]

America’s Poor Get Replaced by Robots

The difference in annual income between households in the top 20 percent and those in the bottom 20 percent -- ballooned by $29,200 to $189,600 between 2010 and 2015. Computers and robots are taking over many types of tasks, shoving aside some workers while boosting the productivity of specialized employees, contributing to the gap. “Technological developments have increasingly replaced low- and mid-skilled jobs while complementing higher-skilled jobs.”

This shift is predicted to continue. About 38 percent of U.S. jobs could be at high risk of automation by the early 2030s. The “most-exposed” industries include retail and wholesale trade, transportation and storage, and manufacturing, with less-educated workers facing the biggest challenges.

Companies’ use of temporary and part-time employees to cut costs also may be widening the disparity, with wage growth failing to keep up with rising residential and basic-necessity expenses. As the divide grows, hardships increase for the bottom 20 percent. Affordable housing, for example, is in short supply nationwide, forcing workers to find shelter further from their jobs and endure lengthier and costlier commutes. Bloomberg

 

[Errol]

On 2017-4-27 at 8:07 AM, rollover said:

About 38 percent of U.S. jobs could be at high risk of automation by the early 2030s. 

 

I think this is an underestimation.

[Errol]

Artificial intelligence to take over half of all jobs in next decade – China's top techie

Robots and artificial intelligence (AI) will replace humans in 50 percent of all jobs in just ten years, says Kai-Fu Lee, founder of venture capital firm Sinovation Ventures and a reputable Chinese technologist.

https://www.rt.com/business/386452-ae-replace-half-jobs-technologist/

[Riedquat]

On 07/04/2017 at 11:27 PM, fru-gal said:

https://futurism.com/this-robot-works-500-faster-than-humans-and-it-puts-thousands-of-jobs-at-risk/

Now show it building something worth having, that makes me look at it and think "This is a nice place to be thanks to that", then I might possibly be won over. Or at least slightly less concerned.

[scepticus]

38 minutes ago, Errol said:

Artificial intelligence to take over half of all jobs in next decade – China's top techie

Robots and artificial intelligence (AI) will replace humans in 50 percent of all jobs in just ten years, says Kai-Fu Lee, founder of venture capital firm Sinovation Ventures and a reputable Chinese technologist.

https://www.rt.com/business/386452-ae-replace-half-jobs-technologist/

When you hear people say stuff like that, its either a sign they know nothing about machine learning or are on the sniff for investment from fools with more money than sense.

In a similar vein, the head of Softbank (the one that recently bought ARM) said recently that in 10 years (or something close to that) our 'sneakers would have more intelligence than us'. Its amazing to me he wasn't laughed out of the room.

[Riedquat]

1 hour ago, scepticus said:

In a similar vein, the head of Softbank (the one that recently bought ARM) said recently that in 10 years (or something close to that) our 'sneakers would have more intelligence than us'. Its amazing to me he wasn't laughed out of the room.

Sad to see ARM go to such a person.

[Sawitcoming]

On 4/26/2017 at 3:26 PM, scepticus said:

The Myth of a Superhuman AI

https://backchannel.com/the-myth-of-a-superhuman-ai-59282b686c62

Interesting and clearly very informed. Perhaps a little too over informed. Although we lack a way to measure intelligence, I do not find it difficult to know that I am smarter than a cockroach or a dog. Of course, if you want to go off into a deep deep analysis of this, the article is probably right. A dog and a cockroach are smarter at some things than me. For all practical and useful purposes in daily human life I am smarter than a dog and a cockroach. 

[Sawitcoming]

On 3/30/2017 at 8:53 PM, Sandwiches33 said:

Yes, I agree and I think bankers and stockbrokers will be the first fully automated.

 Indeed

[Saving For a Space Ship]

Quote

Once a new technology rolls over you, if you’re not part of the steamroller, you’re part of the road.”@stewartbrand 

 

Edited April 30, 2017 by Saving For a Space Ship

[Errol]

"Technology Has Changed The Game": Why The Rise Of Robots Will Be A Permanent Deflationary Force

http://www.zerohedge.com/news/2017-05-02/technology-has-changed-game-why-rise-robots-will-be-significant-deflationary-force

[Riedquat]

On 30/04/2017 at 8:21 AM, Saving For a Space Ship said:

Once a new technology rolls over you, if you’re not part of the steamroller, you’re part of the road.”@stewartbrand 

Yep, that's why it's just depressing and the future looks like being a bleak place indeed.

[Saving For a Space Ship]

Seeking a policy response to the robot takeover

https://www.brookings.edu/opinions/seeking-a-policy-response-to-the-robot-takeover/?utm_medium=social&utm_source=twitter&utm_campaign=es

[wonderpup]

Quote

Interesting and clearly very informed. Perhaps a little too over informed. Although we lack a way to measure intelligence, I do not find it difficult to know that I am smarter than a cockroach or a dog. Of course, if you want to go off into a deep deep analysis of this, the article is probably right. A dog and a cockroach are smarter at some things than me. For all practical and useful purposes in daily human life I am smarter than a dog and a cockroach.

The problem is that 'the myth of a superhuman AI' is itself a myth in terms of the real concerns that people have regarding AI. And -ironicaly- it's people working in AI who seem to be mostly propagating this myth in their efforts to debunk it- they are fighting a straw man largely of their own creation.

Firsty most people have no worries about AI what so ever- they just don't think about it. Among those who do worry about it the main concern is not Superintelligence- or even General intelligence- it's narrow intelligence that just happens to allow a machine or  software to do their job or to reduce the skill needed to do their job to the point where their skillset is made irrelevent in commercial terms.

Conflating some imagined public anxiety re the arrival of superhuman AI with legitimate concerns re the impact of narrow of AI on employment is just muddying the waters of the debate.

In practical terms it does not matter if-for example- a self driving car lacks the smarts to engage in witty banter as it drives it's passengers to their destinations- if I make my living as a taxi driver this lack of social skills on the part of automated vehicles is very unlikely to be a deciding factor if those vehicles offer a cheaper or more convenient service than I do.

So the real threat represented by AI in the near term is not the arrival of General or Superhuman intelligence, but the arrival of artificial idiot savants that do one thing so well that the humans curently employed to do that one thing find themselves surplus to requirements.

Given that automating exisiting jobs is the 'low hanging fruit' when it comes to ROI in AI it's a bit disingenous for the AI 'community' to assert that the real concern is-or should be- the public's alleged anxiety regarding the arrival of Skynet and the Terminator.

Instead of trotting out the dubious claim that the intent is only to 'make people's jobs easier'- rather than replace them- it would be more honest to acknowledge that the entire business model of Artificial Intelligence today is based on the premise that you increase the bottom line by reducing the cost of labour- either by outright replacement with technology or by downskilling the job to point where you can replace expensive skilled labour with cheaper unskilled labour and still get the same results.

While the myth of the the myth of superhuman AI continues to clutter up the debate it's hard to have a serious conversation about the more mundane but far more immediate concerns regarding the impact of AI on our current economic/social arrangements.

Edited May 7, 2017 by wonderpup

[Sawitcoming]

18 hours ago, wonderpup said:

The problem is that 'the myth of a superhuman AI' is itself a myth in terms of the real concerns that people have regarding AI. And -ironicaly- it's people working in AI who seem to be mostly propagating this myth in their efforts to debunk it- they are fighting a straw man largely of their own creation.

Firsty most people have no worries about AI what so ever- they just don't think about it. Among those who do worry about it the main concern is not Superintelligence- or even General intelligence- it's narrow intelligence that just happens to allow a machine or  software to do their job or to reduce the skill needed to do their job to the point where their skillset is made irrelevent in commercial terms.

Conflating some imagined public anxiety re the arrival of superhuman AI with legitimate concerns re the impact of narrow of AI on employment is just muddying the waters of the debate.

In practical terms it does not matter if-for example- a self driving car lacks the smarts to engage in witty banter as it drives it's passengers to their destinations- if I make my living as a taxi driver this lack of social skills on the part of automated vehicles is very unlikely to be a deciding factor if those vehicles offer a cheaper or more convenient service than I do.

So the real threat represented by AI in the near term is not the arrival of General or Superhuman intelligence, but the arrival of artificial idiot savants that do one thing so well that the humans curently employed to do that one thing find themselves surplus to requirements.

Given that automating exisiting jobs is the 'low hanging fruit' when it comes to ROI in AI it's a bit disingenous for the AI 'community' to assert that the real concern is-or should be- the public's alleged anxiety regarding the arrival of Skynet and the Terminator.

Instead of trotting out the dubious claim that the intent is only to 'make people's jobs easier'- rather than replace them- it would be more honest to acknowledge that the entire business model of Artificial Intelligence today is based on the premise that you increase the bottom line by reducing the cost of labour- either by outright replacement with technology or by downskilling the job to point where you can replace expensive skilled labour with cheaper unskilled labour and still get the same results.

While the myth of the the myth of superhuman AI continues to clutter up the debate it's hard to have a serious conversation about the more mundane but far more immediate concerns regarding the impact of AI on our current economic/social arrangements.

Superhuman AI is, AKAIK, a myth currently, this is not in question. The public as you say are not anxious about this. Most of them are unaware of the recent developments. However, for those in the know it is a massively important subject and it is simply a matter of when it will happen, how fast it will happen and what the outcomes might be. 

[Riedquat]

18 hours ago, wonderpup said:

Instead of trotting out the dubious claim that the intent is only to 'make people's jobs easier'- rather than replace them- it would be more honest to acknowledge that the entire business model of Artificial Intelligence today is based on the premise that you increase the bottom line by reducing the cost of labour- either by outright replacement with technology or by downskilling the job to point where you can replace expensive skilled labour with cheaper unskilled labour and still get the same results.

I don't think there's a great deal of thinking about what people actually want to be honest. What's the point of making a job easier? You're getting paid to do whatever it is, easy or hard, so it's no skin off your nose if it's hard. Meaning everyone still has the same skilled work but only needs to work three days a week, everything else stays the same, that would be an improvement, but is that happening?

The real reason there's debate and worry about this whole subject is, as your rightly point out, not about concerns about Skynet or the Matrix. It's because fundamentally it's a development that appears to offer nothing to those who see past the superficial. It's not a technological path that appears to be offering us something that we can't already do. Companies will latch on to it because it'll temporarily get them ahead but in a few years time the playing field will be levelled, so no net gain. It doesn't reduce the consumption of limited, non-replacable resources. It is very much akin to an arms race, nothing more, a damaging vicious circle that's difficult to escape from and benefits none. This is true of a lot of technological "advances", the really useful ones are pretty few and far between.