[NA] Practical tariff impact

The 1 child policy definitely hurt them, but plenty of other industrialized countries out there are also seeing a sustained decline in birth rates.
What makes China hilarious has been the hamfisted approach by the gov't to boost the marriage/family formation rate. They seriously think getting people to have kids is a matter of giving them a check and telling them the gov't needs them to have children. Like the ****ing people are cattle lol.

I was recently reading an article in FT about a company that got an implicit endorsement by the government to implement a policy that would terminate workers who did not get married by a certain date, and also offered company bonuses for people who had kids. And the wording they used sounded like it came right out of Orwell's books: it is your duty to the country.

"Get f***ed" was pretty much the response on Chinese social media. Even the army of censor bots couldn't quell that fire quick enough, so then the government pretended like they were just as outraged as anyone else about the policy :D
 
ArrowFlyer86 said:
The 1 child policy definitely hurt them, but plenty of other industrialized countries out there are also seeing a sustained decline in birth rates.
What makes China hilarious has been the hamfisted approach by the gov't to boost the marriage/family formation rate. They seriously think getting people to have kids is a matter of giving them a check and telling them the gov't needs them to have children. Like the ****ing people are cattle lol.

I was recently reading an article in FT about a company that got an implicit endorsement by the government to implement a policy that would terminate workers who did not get married by a certain date, and also offered company bonuses for people who had kids. And the wording they used sounded like it came right out of Orwell's books: it is your duty to the country.

"Get f***ed" was pretty much the response on Chinese social media. Even the army of censor bots couldn't quell that fire quick enough, so then the government pretended like they were just as outraged as anyone else about the policy :D
Click to expand...

Since you said the O word ...

The entire 20th century was a worldwide struggle between democratic and authoritarian governments in various guises. Fascism, Colonialism, and Communism all fell by the wayside.

That struggle continues in the 21st century. The new incarnation of authoritarianism is perhaps more dangerous because it makes a pretense of democracy. Russia and China appear to be using Orwell as a how-to guide.

I'm old enough to know the world is more complex than it seems and the answers never simple. Like my father, I fought a war premised on the notion of freeing a people who did not fully embrace our vision of freedom.

And yet I think we must continue to stand for a free world, by force if necessary. JFK's inaugural words are as true as when they were said 65 years ago.
 
tspear said:
Yet, only after the one child policy was implemented did the birth rate drop significantly.
Click to expand...
I've yet to see credible evidence that it was actually a policy. I see it mentioned in the USA press. Yet, as @pfarber and I mentioned, we know many people with siblings born during that time. I've looked for citations listing a "one child policy" as some sort regulation or order from their government. I have heard about this policy here, yet I've seen evidence for the opposite there.

It's like lots of people saying that the Great Wall is visible from space, but without evidence. In fact, the Great Wall isn't visible from space with the naked eye.
 
Last edited:
Cap'n Jack said:
I've yet to see credible evidence that it was actually a policy. I see it mentioned in the USA press. Yet, as @pfarber and I mentioned, we know many people with siblings born during that time. I've looked for citations listing a "one child policy" as some sort regulation or order from their government. I have heard about this policy here, yet I've seen evidence for the opposite there.
Click to expand...
I have a friend who is Chinese and now lives in the USA. She tells quite vivid stories from her childhood of people hiding their excess children in walls, and government bulldozers flattening houses with the kids still inside. Girl children left outside to die of exposure because boys are more valuable to a family. Despite these experiences, even while sitting in a restaurant in the USA she was still incredibly reluctant to voice criticism of the Chinese government because she couldn't be sure who might be listening.

Credible? Probably (and understandably) not to you with n=1 and anecdotal. But unless she's the world's greatest actress, she sure as hell convinced me that it was a reality where she grew up.
 
AvNavCom said:
I have a friend who is Chinese and now lives in the USA. She tells quite vivid stories from her childhood of people hiding their excess children in walls, and government bulldozers flattening houses with the kids still inside. Girl children left outside to die of exposure because boys are more valuable to a family. Despite these experiences, even while sitting in a restaurant in the USA she was still incredibly reluctant to voice criticism of the Chinese government because she couldn't be sure who might be listening.

Credible? Probably (and understandably) not to you with n=1 and anecdotal. But unless she's the world's greatest actress, she sure as hell convinced me that it was a reality where she grew up.
Click to expand...

Yeah, "it wasn't actually a policy" is a plot twist and a half. Most people stick with, "it wasn't effective" or "not universally enforced" or something along those lines.
 
After the discussion here, I went and read the wikipedia entry on the policy. It only applied to everyone for a very short period of time, after which it only applied to 35% of the population. So, you're probably all correct.
 
ArrowFlyer86 said:
Like the ****ing people are cattle lol.
Click to expand...
It is the nature of every government to be or try to become that. It always has been, and always will be. It is in the nature of people who aspire to power become that, just as surely as any law of physics. The only fix for this is a relentless effort to fight for individual freedom for everyone. The easy to spot mantra is "we must get people to do this", because the control is overt. But the concept "we must take care of these people", suggestions like UBI, are more dangerous because they are more insidious and will lead to a greater level of control because they weaken rather than strengthen people, and because they attract well meaning and kind-hearted people who don't understand either the damage they are causing, the control they are enabling, or the motives of the backers of the plan.

Or in other words, the first step in having cattle is the plan to feed the cattle.
 
Ed Haywood said:
He says the tariffs are all about China. If we goad them into a trade war and it tanks their growth, it sets back their economic trajectory for overtaking us by about 2 decades. All the other countries are just camouflage for that main objective.
Click to expand...

Ed,

If the Tariffs were all about China, they have done a great job of isolating China.
www.rferl.org

China To Lift EU Sanctions In Move To Repair Ties With Brussels

The discussions to lift the sanctions come as Beijing tries to make inroads against the backdrop of an unfolding trade war with US President Donald Trump’s administration.
www.rferl.org www.rferl.org

Tim
 
tspear said:
Ed,

If the Tariffs were all about China, they have done a great job of isolating China.
www.rferl.org

China To Lift EU Sanctions In Move To Repair Ties With Brussels

The discussions to lift the sanctions come as Beijing tries to make inroads against the backdrop of an unfolding trade war with US President Donald Trump’s administration.
www.rferl.org www.rferl.org

Tim
Click to expand...
Really?

 
Albany Tom said:
But the concept "we must take care of these people", suggestions like UBI, are more dangerous because they are more insidious and will lead to a greater level of control because they weaken rather than strengthen people
Click to expand...
Ok, so how do you think people will get food, shelter and medical care when human labor is no longer needed?
 
Albany Tom said:
It is the nature of every government to be or try to become that. It always has been, and always will be. It is in the nature of people who aspire to power become that, just as surely as any law of physics. The only fix for this is a relentless effort to fight for individual freedom for everyone. The easy to spot mantra is "we must get people to do this", because the control is overt. But the concept "we must take care of these people", suggestions like UBI, are more dangerous because they are more insidious and will lead to a greater level of control because they weaken rather than strengthen people, and because they attract well meaning and kind-hearted people who don't understand either the damage they are causing, the control they are enabling, or the motives of the backers of the plan.

Or in other words, the first step in having cattle is the plan to feed the cattle.
Click to expand...
Truth.

This is why I always feel pity towards the ignorant that pine for a distributed wealth utopia like Gene dreamt of in Star Trek.
 
Albany Tom said:
It is the nature of every government to be or try to become that. It always has been, and always will be. It is in the nature of people who aspire to power become that, just as surely as any law of physics. The only fix for this is a relentless effort to fight for individual freedom for everyone. The easy to spot mantra is "we must get people to do this", because the control is overt. But the concept "we must take care of these people", suggestions like UBI, are more dangerous because they are more insidious and will lead to a greater level of control because they weaken rather than strengthen people, and because they attract well meaning and kind-hearted people who don't understand either the damage they are causing, the control they are enabling, or the motives of the backers of the plan.

Or in other words, the first step in having cattle is the plan to feed the cattle.
Click to expand...
Never attribute to malice what can be adequately explained by incompetence.
 
Tarheelpilot said:
This is why I always feel pity towards the ignorant that pine for a distributed wealth utopia like Gene dreamt of in Star Trek.
Click to expand...
Unless you are going to reject the hypothesis that human labor will soon no longer be necessary, you have to explain how wealth will be distributed when human labor is no longer necessary. So, how do you think that will happen if it isn't via a distributed wealth utopia like star trek?
 
Tarheelpilot said:
I’m really surprised this thread is still open. Seems like we need to reopen the spin zone.
Click to expand...
While there are a few posts that toe the line, it seems like most of the responses are more intellectually-driven rather than political. It's going much better than I figured it would, lol.
 
172andyou said:
Unless you are going to reject the hypothesis that human labor will soon no longer be necessary,
Click to expand...
This is a “hypothesis” without any backing in fact.

This concept has been aired in hushed voices trembling with fear for centuries - every time a new disruptive technology came around.

It’s never happened before, and it won’t happen now. The ways in which people work will change, certainly - but it won’t vanish, any more than it vanished when the cotton gin was invented, tractors replaced horses, the Xerox machine was invented, the calculator became a reality, nor when the computer became ubiquitous in business decision-making.

It’s nothing more than fear-speech borne out of ignorance and a lack of vision of what the future can hold for our race. It is then supported by politicians with a desire to use that fear to maintain and increase their level of influance and control.
 
StraightnLevel said:
This is a “hypothesis” without any backing in fact.
Click to expand...
Other than the current existence of machines that have already automated people out of jobs and AI that is better at some jobs than humans. Today. Not hypothetically, today.

StraightnLevel said:
This concept has been aired in hushed voices trembling with fear for centuries - every time a new disruptive technology came around.

It’s never happened before, and it won’t happen now. The ways in which people work will change, certainly - but it won’t vanish, any more than it vanished when the cotton gin was invented, tractors replaced horses, the Xerox machine was invented, the calculator became a reality, nor when the computer became ubiquitous in business decision-making.

It’s nothing more than fear-speech borne out of ignorance and a lack of vision of what the future can hold for our race. It is then supported by politicians with a desire to use that fear to maintain and increase their level of influance[sic] and control.
Click to expand...
Got it. So your bet is that, despite the first instances already being in production today, you're betting that somehow progress will just stop and humans will remain necessary.

I'd bet you, but I'm not going to be around long enough to collect the money on a hundred year wager. Which is a shame since it seems so obvious it's going to happen. :-)
 
172andyou said:
Other than the current existence of machines that have already automated people out of jobs and AI that is better at some jobs than humans. Today. Not hypothetically, today.


Got it. So your bet is that, despite the first instances already being in production today, you're betting that somehow progress will just stop and humans will remain necessary.

I'd bet you, but I'm not going to be around long enough to collect the money on a hundred year wager. Which is a shame since it seems so obvious it's going to happen. :-)
Click to expand...
CNCs didn’t obsolete machinists.

Calculators didn’t obsolete engineers nor accountants.

Personal computers haven’t obsoleted much of anyone, except perhaps mailmen and people who wrote chain letters.

Every time a new tech comes along that reduces labor input in some way, that labor gets redirected to make use of that tech to generate greater value.

Like I said, the “hypothesis” is fear-speech borne out of ignorance and lack of vision.
 
StraightnLevel said:
CNCs didn’t obsolete machinists.

Calculators didn’t obsolete engineers nor accountants.

Personal computers haven’t obsoleted much of anyone, except perhaps mailmen and people who wrote chain letters.

Every time a new tech comes along that reduces labor input in some way, that labor gets redirected to make use of that tech to generate greater value.

Like I said, the “hypothesis” is fear-speech borne out of ignorance and lack of vision.
Click to expand...
Yeah, yeah, you can keep calling me/us ignorant and visionless. Sticks and stones.

But you still fail to address the reality that the machines will be faster, cheaper and more capable than humans sometime soon. When that happens why would I, as a business owner, ever hire a human again? That is a transition that has never happened before, so your examples are misplaced. There is no place to redirect human labor to because they would be competing against a machine that is faster, cheaper and more capable.

Or, to flip it around, what's your vision for what a human will do that a machine won't be able to do faster, better and cheaper? What are the jobs that I, as a business owner, would be hiring people to work because there is no way to make a machine that is faster, cheaper and more capable?
 
172andyou said:
Other than the current existence of machines that have already automated people out of jobs and AI that is better at some jobs than humans. Today. Not hypothetically, today.
Click to expand...

AI is basically a parlor trick. Nothing more than probability analysis on large data sets. AI is still 100% dependent on people to frame each problem so the machine can respond with a collection of yes/no answers.

AI is also flummoxed by simple and obvious tricks that any human sees through immediately. The Tesla painted road test is a recent example.

Despite decades of effort, nobody has produced a robot capable of human mobility on other than a closed course.

Nobody on this board will see the elimination of human labor in their lifetime.
 
Ed Haywood said:
AI is basically a parlor trick. Nothing more than probability analysis on large data sets. AI is still 100% dependent on people to frame each problem so the machine can respond with a collection of yes/no answers.
Click to expand...
You are describing the current LLM fad and thinking it represents all of AI. It doesn't.

You are also assuming that the rapid progress of the past five years or so won't continue. I think it will.

Ed Haywood said:
AI is also flummoxed by simple and obvious tricks that any human sees through immediately. The Tesla painted road test is a recent example.
Click to expand...
1) The entire point of that video was that using lidar could improve capability. They illustrated that case by using a car that is shipping today. Not forward looking speculation, but something shipping already. So not a very good example to use to make the case that AI will always be easily flummoxed.

2) Further, didn't someone test with the latest Tesla hardware and it passed? That's how fast things are moving.

Ed Haywood said:
Despite decades of effort, nobody has produced a robot capable of human mobility on other than a closed course.
Click to expand...
1) Boston Dynamics is making huge progress every year.

2) Not even sure 'human mobility' is the right benchmark. I would bet that the future looks more like a bunch of different form factors specialized to tasks, rather than human mobility. See my previous note about plumbing repair. Duplicating human mobility for repairing plumbing would be counter productive. There are real liabilities to being human and trying to do plumbing in small, enclosed places for example.
Ed Haywood said:
Nobody on this board will see the elimination of human labor in their lifetime.
Click to expand...
While the demo here skews older, we do also have some 20-somethings. I bet they witness the vast majority of this transition before they shuffle off this mortal coil. 60-70 years is a very long time at the pace technology is changing. Factories today are almost as different to those in 1955 as those in 1955 were to those in 1885.
 
172andyou said:
You are describing the current LLM fad and thinking it represents all of AI. It doesn't.
Click to expand...

No, I am not. I am describing the general function of neural networks, which is the foundation of LLMs, image recognition, and most other AI developments of the last 2 decades. I am quite familiar with the underlying algorithms and tools of AI.
 
172andyou said:
Or, to flip it around, what's your vision for what a human will do that a machine won't be able to do faster, better and cheaper? What are the jobs that I, as a business owner, would be hiring people to work because there is no way to make a machine that is faster, cheaper and more capable?
Click to expand...
So, you still think we’ll need the business owner. That’s a great starting point.

Why does the business need a human owner?
 
172andyou said:
Unless you are going to reject the hypothesis that human labor will soon no longer be necessary,
Click to expand...
While labor requirements will be reduced on the production output side, the labor requirements, ie, people to maintain the automation infrastructure, will go up as is happening today. So it will be more a shift in labor skill sets as automation, AI, etc. increase and enter new industries. And when the numbers are plugged in for the growth in automation, they show a shortage of trained technicians/mechanics to maintain those machines. Unfortunately, given the skill set an A&P comes out of school with, they are becoming a sought-after commodity within automation industry just as petroleum directional drilling companies and dental equipment OEMs did in the 80s.
 
Ed Haywood said:
No, I am not. I am describing the general function of neural networks, which is the foundation of LLMs, image recognition, and most other AI developments of the last 2 decades. I am quite familiar with the underlying algorithms and tools of AI.
Click to expand...
Neural networks are a popular approach, but there are a dozen others. As I guess you know given your familiarity.

Humans didn't lose to machines with chess playing neural networks, for example. Deep Blue used a different approach. Expect that to continue. Different algorithmic approaches for different classes of problems.

And, in any case, be careful claiming neural networks as "nothing more than probability analysis on large data sets" since we are also neural networks. Whiskey and a campfire would be a more appropriate setting for the discussion of whether humans are doing anything more than probability analysis on large data sets.

StraightnLevel said:
So, you still think we’ll need the business owner. That’s a great starting point.

Why does the business need a human owner?
Click to expand...
As I wrote a bunch of messages before, one of the only roles that seem likely to continue will basically be oversight of the machines. I don't think we need the business owner from the standpoint of the human providing something that the machine can't do. I think we will refuse to turn the planet over to them without oversight.

Bell206 said:
While labor requirements will be reduced on the production output side, the labor requirements, ie, people to maintain the automation infrastructure, will go up as is happening today.
Click to expand...
I see no reason to believe machines won't be able to do that.
 
My experience is a couple of years old so things may have advanced. Deep neural networks do some pretty amazing things making non-obvious connections. But two things really limited their usefulness when I was actively using them: they require a great deal of data (carefully curated data) to learn and make those connections, someone has to tell them what criteria to use to determine “goodness”. I don’t see general purpose AI taking over things that don’t have lots of data to train them. (By lots I’m talking multiple millions of pieces of data.)
Does anybody here know if that amount of data to train has changed significantly?
 
jsstevens said:
My experience is a couple of years old so things may have advanced. Deep neural networks do some pretty amazing things making non-obvious connections. But two things really limited their usefulness when I was actively using them: they require a great deal of data (carefully curated data) to learn and make those connections, someone has to tell them what criteria to use to determine “goodness”. I don’t see general purpose AI taking over things that don’t have lots of data to train them. (By lots I’m talking multiple millions of pieces of data.)
Does anybody here know if that amount of data to train has changed significantly?
Click to expand...
That's an interesting question. Are you talking about the underlying model, or about the domain-specific training? The former likely will always require a huge data set. The latter really depends on the use case and the recall/precision requirements of the customer. Use cases and consequences of incorrect responses can vary widely. Think corporate knowledge stores/help desks as opposed to cancer detection, for example.
 
jsstevens said:
My experience is a couple of years old so things may have advanced. Deep neural networks do some pretty amazing things making non-obvious connections. But two things really limited their usefulness when I was actively using them: they require a great deal of data (carefully curated data) to learn and make those connections, someone has to tell them what criteria to use to determine “goodness”.
Click to expand...
The current LLMs don't use "good" as much as "connected". Good is a value statement. Connected is objective. But all the current stuff uses multiple approaches anyway.

"Good" will always have a role. For example, machines that design and construct buildings have to obey physics and governmental laws. So those will be used not just to say "good" but "required" and models trained to require adherence.

jsstevens said:
I don’t see general purpose AI taking over things that don’t have lots of data to train them. (By lots I’m talking multiple millions of pieces of data.)
Does anybody here know if that amount of data to train has changed significantly?
Click to expand...
The big models, like ChatGPT, use about 2 trillion pieces of data. There were very significant improvements in capability with each order of magnitude going from thousands to millions to billions. But it is much less clear that going from mid-billions to trillions was similarly helpful. So there may be an upper bound beyond which adding more training data doesn't move the needle much. For the past year or two, some of the bigger advancements have been in model tuning rather than increases in training data volume.
 
Baked Potato said:
That's an interesting question. Are you talking about the underlying model, or about the domain-specific training? The former likely will always require a huge data set. The latter really depends on the use case and the recall/precision requirements of the customer. Use cases and consequences of incorrect responses can vary widely. Think corporate knowledge stores/help desks as opposed to cancer detection, for example.
Click to expand...
I was talking about underlying models. We used googles pretrained image recognition and did domain specific training to pick out what we were after. Took several hundred images rather than many millions.
 
172andyou said:
The current LLMs don't use "good" as much as "connected". Good is a value statement. Connected is objective. But all the current stuff uses multiple approaches anyway.

"Good" will always have a role. For example, machines that design and construct buildings have to obey physics and governmental laws. So those will be used not just to say "good" but "required" and models trained to require adherence.


The big models, like ChatGPT, use about 2 trillion pieces of data. There were very significant improvements in capability with each order of magnitude going from thousands to millions to billions. But it is much less clear that going from mid-billions to trillions was similarly helpful. So there may be an upper bound beyond which adding more training data doesn't move the needle much. For the past year or two, some of the bigger advancements have been in model tuning rather than increases in training data volume.
Click to expand...
Raises an interesting question in my mind. ChatGPT was fed “the internet “ as training data. But today more and more of “the internet “ is machine generated by ChatGPT and its ilk. Do they start to get distorted views based on circular training?
 
jsstevens said:
Raises an interesting question in my mind. ChatGPT was fed “the internet “ as training data. But today more and more of “the internet “ is machine generated by ChatGPT and its ilk. Do they start to get distorted views based on circular training?
Click to expand...
ChatGPT used a wide variety of data, including literal libraries of books, newspaper and other more well regarded publishing than PoA. :-)

But, yes, recursion is an area of concern in the industry. And, yes, crazy things can happen.
 
172andyou said:
[snip] :-)

But, yes, recursion is an area of concern in the industry. And, yes, crazy things can happen.
Click to expand...
After 40+ years in the software business, that is soooo true!
 
172andyou said:
I see no reason to believe machines won't be able to do that.
Click to expand...
I don’t believe so. 30 years ago I was told computers and automation would reduce my maintenance workload and make some of my requirements redundant. Instead, it only changed how I approached a task and created new problems to deal with. And from what I’ve seen in the current cutting-edge side of non-aviation automation it will be just more of the same for the next 30 years: a shift in how that equipment is serviced and maintained by those steadfast carbon-units.
 
172andyou said:
You are describing the current LLM fad and thinking it represents all of AI. It doesn't.

You are also assuming that the rapid progress of the past five years or so won't continue. I think it will.


1) The entire point of that video was that using lidar could improve capability. They illustrated that case by using a car that is shipping today. Not forward looking speculation, but something shipping already. So not a very good example to use to make the case that AI will always be easily flummoxed.

2) Further, didn't someone test with the latest Tesla hardware and it passed? That's how fast things are moving.


1) Boston Dynamics is making huge progress every year.

2) Not even sure 'human mobility' is the right benchmark. I would bet that the future looks more like a bunch of different form factors specialized to tasks, rather than human mobility. See my previous note about plumbing repair. Duplicating human mobility for repairing plumbing would be counter productive. There are real liabilities to being human and trying to do plumbing in small, enclosed places for example.

While the demo here skews older, we do also have some 20-somethings. I bet they witness the vast majority of this transition before they shuffle off this mortal coil. 60-70 years is a very long time at the pace technology is changing. Factories today are almost as different to those in 1955 as those in 1955 were to those in 1885.
Click to expand...
Using your plumbing example, I have a really hard time believing that sending a multitude of automated machinery to a house in order to replace a water heater is going to be more efficient/effective than using a traditional plumber. The problem with stuff that's relatively simple (like most plumbing) is that it's rarely the same in every instance. One house uses a water heater in the garage with standard fittings, another has one in the attic space above the garage and has all manner of oddball connections/adapters. Cast iron to PVC, copper sweating, compression fittings, etc. just end up having so many variables that a human can easily adapt to which a specialized machine can't accomplish. Not to mention being able to lift/remove/reinstall the water heaters.

While we can all appreciate the thought of an iRobot future, I don't think we're anywhere near that point in the human timeline. The automation works better in systems that are predictable and can have fairly well-controlled inputs. I can see many grocery stores going more automated (like an Amazon warehouse) and your weekly grocery order showing up at your door without a human lifting a finger. Local transit systems with predetermined stops and routes can be done with automation with potential to use AI to handle emergencies. Things like teaching school children, medicine, or any sort of construction related to repair/remodel is going to be a long way (over a century) off due to the high variability in those environments, in my opinion. I can be optimistic about the advances in technology we've had in the past century while acknowledging that we are still a long ways away from needing UBI or having enough displaced workers for it needing to be worked into our economic way of life.

As someone who has worked in manufacturing for 20 years, I think you'd be really surprised at how similar most facilities are to the early 1900s. Still lots of manual labor, people sitting at stations or assembly lines doing trimming with a razor blade or grinder, applying adhesive, etc. Some things can be automated easily when you're running large batch sizes. Running 100 different SKUs per month in low quantities doesn't really lend itself to the purchase, design/layout of a production floor, and writing of programs to try and make automation work. I'd pose that you can often run a factory for a decade on the labor cost that it would take to automate everything. Not many owner/operators are going to commit that kind of up-front capital to the automation, much less have the collateral to ensure the company remains a going-concern on that scale of timeline.
 
SoonerAviator said:
Using your plumbing example, I have a really hard time believing that sending a multitude of automated machinery to a house in order to replace a water heater is going to be more efficient/effective than using a traditional plumber. The problem with stuff that's relatively simple (like most plumbing) is that it's rarely the same in every instance. One house uses a water heater in the garage with standard fittings, another has one in the attic space above the garage and has all manner of oddball connections/adapters. Cast iron to PVC, copper sweating, compression fittings, etc. just end up having so many variables that a human can easily adapt to which a specialized machine can't accomplish. Not to mention being able to lift/remove/reinstall the water heaters.

While we can all appreciate the thought of an iRobot future, I don't think we're anywhere near that point in the human timeline. The automation works better in systems that are predictable and can have fairly well-controlled inputs. I can see many grocery stores going more automated (like an Amazon warehouse) and your weekly grocery order showing up at your door without a human lifting a finger. Local transit systems with predetermined stops and routes can be done with automation with potential to use AI to handle emergencies. Things like teaching school children, medicine, or any sort of construction related to repair/remodel is going to be a long way (over a century) off due to the high variability in those environments, in my opinion. I can be optimistic about the advances in technology we've had in the past century while acknowledging that we are still a long ways away from needing UBI or having enough displaced workers for it needing to be worked into our economic way of life.

As someone who has worked in manufacturing for 20 years, I think you'd be really surprised at how similar most facilities are to the early 1900s. Still lots of manual labor, people sitting at stations or assembly lines doing trimming with a razor blade or grinder, applying adhesive, etc. Some things can be automated easily when you're running large batch sizes. Running 100 different SKUs per month in low quantities doesn't really lend itself to the purchase, design/layout of a production floor, and writing of programs to try and make automation work. I'd pose that you can often run a factory for a decade on the labor cost that it would take to automate everything. Not many owner/operators are going to commit that kind of up-front capital to the automation, much less have the collateral to ensure the company remains a going-concern on that scale of timeline.
Click to expand...

Variability is kryptonite to any automation solution design.
 
CJones said:
Variability is kryptonite to any automation solution design.
Click to expand...
Agreed, and as @jsstevens mentioned regarding neural networks (and other AI), it takes millions upon millions of carefully curated data points in order to get automated responses to line up with changing variables for even somewhat typical tasks for a human. Since no two factories are 100% identical, it makes replication of those automations incredibly difficult on a large scale. It's just a mountain of a task to try an implement even for basic stuff, so we're still figuratively at basecamp of Mount Everest at this point. Long way to go in order to build enough cognitive learning into the software to adapt to so many planned and unplanned variables.
 
172andyou said:
Unless you are going to reject the hypothesis that human labor will soon no longer be necessary,...
Click to expand...

Yes, I reject it. Similar hypotheses were proposed all through the '50s and '60s regarding computers. Didn't happen then, won't happen now, won't happen a hundred years from now. Every time some technical advancement comes along, prophets of doom forecast tragic job losses, yet the unemployment rate remains fairly constant.

Humans have an innate need to work, to create, to feel useful, to contribute to their own well being, to care for their families, and to improve society. That won't change. As automation takes over some tasks, humans will be freed to pursue other interests, and that will include work and the advancement of civilization.


172andyou said:
What are the jobs that I, as a business owner, would be hiring people to work because there is no way to make a machine that is faster, cheaper and more capable?
Click to expand...

Well, here's a short starter list. Remember, per your rules, we're talking replacing people with automation, not people using automation to accomplish a job.

- Overhauling my Lycoming O-360.​
- Shoeing my wife's horse.​
- Training said horse to walk onto a trailer and jump over a fence.​
- Replacing my roof.​
- Installing or repairing our pasture fencing.​
- Finding and arresting criminals.​
- Dentistry.​
- Surgery.​
- Gunsmith. (I can't imagine any machine repairing and glassing the stock on my 101 year old LC Smith.)​
- Fishing guide.​
- Hunting guide.​
- Ballroom dancing instructor.​
- Installing a custom rollcage into my racecar.​
- Children's swimming instructor.​
- Scuba diving instructor.​
- Topless waitress.​
- Martial arts instructor.​
- Hair stylist.​
- Carpet installer.​

I'd go on but my fingers are getting tired.
 
Half Fast- one of your examples was rather intriguing…..
 
You must log in or register to reply here.
Back
Top