Of course, but then you are ignoring how technology in the future will likely help that bottom 25%; but once again, IQ is not destiny, except for the number of who are really at the bottom, you should be talking about 14% to under 5% regarding who can be considered to be at un-trainable (for the time being) levels.
Sure they are, becuase to be honest, food service doesn’t require all that much intellectual rigor, and those with more talents in the academic intelligence department are going to be putting those talents towards more interesting things.
And as far as labor saving devices, you cannot use a labor saving device just because you have researched it and you think that it saves you labor, you can only use the devices that the rating agency allows you to use. It was fairly recent that using electrically powered anything was allowed if you wanted a 5 star (3 in michelin) rating. Some culinary schools still do not allow electric mixers or anything else.
Top 75% in intelligence - however you measure it - is hardly genius level. Top 50% could do these things no problem - maybe top 60%. And I’m talking about creating, not implementing.
I’m sure you are right about top restaurants, but automation of cooking in chain restaurants is going to happen for cost reasons, not because they can’t find someone smart enough to do the cooking.
For cooking at home, no correlation at all with intelligence. I know someone who got 1600 on her SATs and has a PhD who either can’t cook or won’t cook.
If you give someone with an atrophied arm a new one, and they can now pitch in the big leagues, you can’t really claim that people with atrophied arms can be good pitchers.
And if this technology enhances the brain for everyone (I’m far from convinced) what will happen is that those on top will get it first and the gap will become even wider. You might pay to try this on one Charley Gordon, but on all of them?
Again, it’s a huge mistake to imagine that it will be all the dumb people who won’t have a job. No, the current wave of automation is going to overwhelmingly replace so-called skilled workers. Or to put it a slightly different way, improvements in tools are going to make skilled workers vastly more productive, which means a collapse in the job market for people in those professions.
The last jobs to be automated out of existence will be the ones that just need a human being to use their normal eyes and ears and hands and legs, not the ones that use a skilled human brain.
And again, “robots” and “AI” aren’t the main way jobs get eliminated/productivity is increased. It’s replacing labor-intensive ways of doing things with non-labor intensive methods. Nobody is going to invent a bricklaying robot that could replace a skilled mason. Instead they’ll use some other building technique that gets rid of bricklaying entirely. Except when you want that real brick for show, even though bricks cost ten times as much and function worse than some future extruded building product.
Or take journalism. Despite everyone’s best efforts, we still don’t have computer systems writing news stories. But jobs in journalism have collapsed anyway. And that’s simply because you don’t need 1000 people from 1000 newspapers writing the same story about the president’s speech today. Instead you have one guy write the story once, and then the 1000 newspapers–or rather 100,000 websites–repost or link the story.
For every creator, you have several implementers.
Oh, yeah, burger flipping is going to be done by flippy the robot, not phil the cook.
And one of the dumbest people I know makes some of the best home cooked meals I’ve ever had.
Right, technology may not take your job, but it will let you do the work of 10 other people, which means that we only need 10% of your job’s workforce now.
Dog grooming FTW!
I think that that job is actually more important, and has gone by the wayside not because of increased supply, but because of decreased demand, specifically, demand that is willing to actually pay. I know I am guilty of this myself, but I cannot afford subscriptions to the dozen papers you need to read in order to keep up with things. There is no robot that has taken the place of a copy editor, and it shows.
Which is exactly what has been happening for the last 200+ years…and accelerating. What you aren’t mentioning is that freeing up that labor will free it up to do something else…as has happened repeatedly when this has happened in the past. I’ve seen estimates that up to 30% of the current workforce will almost certainly be replaced due to automation in the next 20 years, and I believe it. But that’s not unusual either…we’ve seen similar displacements in the past, some even more dramatic such as the farm sector example used earlier.
I think the real difference this time is the AI angle, not the automation or straight productivity. But I see the AI thing as having the potential to actually make human work more valuable, since fusions between humans, machine and AI seem to be the optimal mix for a lot of jobs, and I think that trend will explode in the next half century. Or we will hit the singularity where there seems to be a binary result set…either we, as a species will live forever, or we will become extinct.
We are all living in really interesting times. 
The take away here should be that attempting to predict the future is a fools game. Look back at all of the past predictions of gloom and doom when societies were faced with similar things to what we are discussion right now. Those predictions were almost uniformly bad and way off the mark. Today, less people world wide live in poverty than at any time in history, at the same time where we have an increasing number of people…predictions of both in the past were that poverty would be worse and populations would starve at a billion people.
I don’t think that automation is a bad thing, but I do think that it will cause displacement. People who have jobs now will find those jobs no longer available in fairly short order. As you said, the pace is increasing, so a larger number of people will be displaced and looking to fill the jobs that are left, and the training for the new jobs created will have lag time, and not everyone will be able to fulfill the training.
Like I said, it makes people more productive, or “valuable”, but that means that it takes fewer to produce the same goods and services. Unless we consume more, and have the income to do so, then there will be fewer jobs needed to fulfill the needs and desires of the populace.
I don’t look forward to the future with doom and gloom, but I do look at it with a bit of apprehension that we can easily take a path that does lead to great misery, or we can take a path that lead us to a “utopia”, and I don’t know if we will know what path we are on until it is too late to change. This is compounded by the fact that those who are going to be choosing the path are not the ones that will suffer the consequences if it is a poor path to follow.
And I did not say that, (knowing how MLB reacts against performance enhanced drugs, performance enhancement bionics will be out) but for switching among tasks that are not so simple, it will do.
As the people that can not be trained for having a handicap are not hard to identify, it is more likely that automation will reduce the prices and increase the availability of technological solutions for the people that are affected.
I do foresee medicine as one of the big reasons why AI is being developed, after all that is the main reason why Watson is being developed. Yes, developed, even after winning Jeopardy there are still more things that it needs to learn before becoming effective in medicine, the point here is that there re still human doctors that have to intervene. But the promise is that a lot of the problems that we want to solve thanks to automation and the growth of AI are also related to the people that are in need.
A certain amount of healthy caution is never a bad thing, and in this case I also have equal measures of hope and apprehension wrt how things will pan out. I do think that a lot of different things are converging at this time, and that our society is on the cusp of true greatness or some really bad times. My own optimism points more towards looking at the positive things we as a species have already done, and that emergent AI could help us to do all the things our potential warrants, but I’m realist enough to know that we could fuck it all up right at the end.
I think this is the wrong way of looking at it. It will take fewer people to produce the goods and services that we use today, but that doesn’t mean we will need fewer workers in the future. Just like today we produce an abundance of food with fewer and fewer workers, but those workers aren’t all sitting around on unemployment, there are things we do that were unimaginable to the people at the turn of the previous century. And I think that will be the case down the road too. Things we haven’t even thought of, or don’t really consider ‘work’ today will suck up that labor, IMHO. What it will be I have some ideas, but that probably means they are wrong and what will actually happen is something no one today could predict.
I will say that I think people in this thread are under rating the impact of entertainment in the future. I was watching a crowdfunded movie trailer the other day by a couple of guys who basically did it in their basement, and it was pretty good. The tools available today make it possible for a few people to do things that it took large studios and an entertainment industry to do in the past. Consider the impact if a handful of motivated people using some near future tools could put together a movie or other entertainment program without a big budget, studios and bean counters getting in the way, etc etc. There is an emerging world wide market out there…literally billions of connected people, many with the means of buying and consuming entertainment, and who might not necessarily need their entertainment to come from some Hollywood studio. And that’s not even the tip of the very large iceberg. Think of all the things those billions of users might want to consume in the future. Then think of the relative handful of movie studios out there using the same old formulas to churn out the same old movies, often rehashed because some bean counter or big wig figured, hey, it made money before, why not again? Eventually, I see that as I see the old broadcast TV channels, replaced by cable networks, being replaced as we speak by on demand content providers who are increasingly doing their own content. And what else might those billions want or need? Look at things like Twitch or YouTube channels.
Absolutely there will be people displaced. And this time, it might not just be blue collar workers. You could see doctors and lawyers being displaced, or other white collar workers. And it’s possible we will see an entire sector downsized, as I think we are on the cusp of autonomous driving capabilities. Couple that with automated warehouses and the entire logistics sector could be disrupted. And a lot of those folks will get caught in the gears of progress. We will need, more than ever, to have capabilities and resources available to help those displaced find new things to do, new training. I think we, as a society should seriously look at ‘free’ after high school training and schooling. It’s what we as a society did in the farm displacement in the past, and it gave us incredible returns on our investment. I think this will as well, allowing us to train and educate a workforce for the new era. We just have to get past some of the stupid politics and attitudes stopping this, but I think we need to do it by explaining to the people WHY we are doing this, not just saying ‘free education for all because it’s a right!’. Just like when you have to sell something to upper management we need to sell these sorts of ideas to the public by explaining the cost to benefit and WHY we are doing it…and what the return on investment will be in tangible terms.
I do wonder if any of that will change. I do actually look forward to the day when we have genetically and/or cybernetically enhanced athletes battling it out at performance levels that make current olympians look like clumsy puppies.
I do think that if we solve the healthcare crisis (and to be honest, single payer or other forms of UHC only help, not solve the problem), it is going to be with substantial assistance by computers and robots. If you can answer detailed surveys and history with the computer, and it collates that info with test results, that can give the human doctor quite a bit of assistance in diagnosing and treating whatever it is that ails you. Most doctor visits could be handled by a nurse practitioner, or even really just a nurse for most things, freeing up the doctor to be involved in more complicated cases that require more human interaction.
They can also be invaluable in researching of new cures and treatments.
Short of some sort of catastrophe, I think that life will be great for at least 10% of the population. Whether life is great for the other 90% is going to be largely up to the first 10%. The resources will be available, but whether they will be allocated is another matter.
The change is happening more rapidly than it has in the past, and even with the rates of change in the past, many workers were left displaced with little to nothing productive to do. They need training and relocation, both of which not only consume resources, but also get much resistance from people that don’t want to change.
Some coal miners are going on to training to be web developers or computer programmers, but, due to both lack of adequate resources to train everyone, and lack of desire on the part of the displaced workers to be retrained, most are staying in coal country, collecting welfare until their coal mining job comes back.
Whether or not the end result is where everyone is as productive as they care to be to receive the goods and services they desire or not, the transition is likely to be at least a bit rough under the best of circumstances, and can be nearly catastrophic if it is not carried out smoothly.
Entertainment is going to probably be the employment of most people looking to have a job. The only problem is is who pays for it? Right now, we have jobs doing all sorts of things, and then we give a bit of that money to the entertainment industry to pay for them making movies that we want to watch. We let people try to sell us things that we may have some intention and capability of purchasing in exchange for television shows. If most people don’t actually have an income to spend on either entertainment or on advertised products, both of those forms of paying for entertainment dry up.
It is of course in the best interest of society to make its members as productive as possible. Relying upon them to use their own resources to attain the skills and knowledge required to be productive seems counterintuitive, especially as the wealth that is in the hands of individuals dwindles. Part of a comprehensive safety net that includes food, clothing, shelter, and healthcare should also e education. And computers and robots can help there. I am sure you have seen the youtubers doing product endorsement for brilliant or curiosity stream, and while I have yet to check any of those out myself, I assume that they are relatively useful in their educational resources offered. Making such systems more robust and offering them for free to the public could go a long way towards closing education gaps. I don’t think that automation can completely replace teaching, at least not in the near future, but automation and machine learning can also free people up to be more individual tutors than classroom teachers.
In this thread, we have mostly discussed direct economic impacts of AI/automation, and then the impacts this disruption might have on the political sphere. But this NY Times op-ed focuses on the direct threat AI poses to politics and democracy itself, in the form of chatbots:
Hello, 2nd post on the dope. I’ve really enjoyed reading this thread even though I’m only about halfway through it.
I just wanted to comment about a post I read back quite a few pages about lights out manufacturing. I work in manufacturing and wanted to comment.
I visited a plant in japan back in 2001 that practiced lights out. But here’s the thing, this part of the plant machined huge castings for machine tools. The cycle time for machining one of these castings was ~22 hours. So, the first shift would setup the operation and then using their cnc mills, would then let them run all night. Wasn’t that impressive, just the nature of the operation. If the cycle time were only 8 hours, they’d keep the lights on longer. No AI involved, just CNC’s that have been around for 30 years+
Interesting. And welcome to the Dope! I can imagine how intrigued I would be if I just found this thread. But it’s a lot by this point! Kudos for reading so much already.
Thanks SlackerInc! It’s a pleasure to read such a thought provoking thread. I work for a moderately sized global japaneses manufacturer, who have been using automation in its factories since the ‘80’s, like many Japanese manufacturers. The Japanese have been on the for front of adopting automation and industrial robots. IME what we see in the US as far as factory automation, has already occurred in Japan, 30 years ago. What we’re seeing now in the US is just “catch up”, due to automation costs finally becoming competitive here in the US. But this is just the easy automation, the repetitive tasks, just like japan 30 years ago. I have worked for a heavily automated factory for the last 20 years, and all the easy repetitive tasks were automated 30 years ago, there hasn’t been much change since then and haven’t seen much improvement in automation that would eliminate many jobs from the trade shows I attend.
I guess what I’m trying to convey is that the automation of the last 30 years has been relatively cheap and easy to reduce labor by say ~50% (just an example), but it will become increasingly difficult and more expensive for each percentage moving forward. Hope this makes since, I’m not the best writer.
I think I’ve got the gist. Basically, you think the low-hanging fruit was grabbed thirty years ago in Japan, and now the U.S. is catching up, but we should then hit a kind of plateau. Is that about right?
It makes a certain amount of sense. I wonder, though: what about the massive increases in computing power in recent years and developments in AI? You’ve got outfits like Boston Dynamics making strides unimaginable even one decade ago, much less three. You don’t think that will filter its way into the manufacturing or service sectors?
Yes! That’s exactly what I was trying to say, thanks for deciphering.
I’m sure there’s plenty of applications for the Boston Dynamics robots, but might be overkill for automotive production. Industrial robots usually perform repetitive tasks where speed and accuracy are required. An actual human looking robot isn’t really necessary if wheels can work just as well and be faster and more accurate. But I’m sure there are many applications out side of the automobile manufacturing sector.
IMO, the next wave of labor reduction in the automotive manufacturing sector will happen with the shift to EV's, since EV's require far fewer and simpler components.We’ve been having the debate about ‘robots taking all the jobs’ since the time I started on the SDMB 19 years ago. In one of those early threads I mentioned that the debate was not new, and I remember people confidently asserting that robots would take all the jobs ‘within ten or twenty years’ back in college in the 1980’s. Like fusion power, automation of all jobs is only ten or twenty years away - and has been for forty years.
So here we are, decades later, and unemployment in the U.S. is at an all-time low. And I predict that two decades from now, whatever the unemployment rate is will have everything to do with overall economic conditions, and almost nothing to do with automation - just like it is today. And the reasons are still the same - humans have general intelligence, which is way more valuable than most people in the debate seem to recognize, and replacing that makes automation much harder than people seem to realize, except in those areas where general intelligence doesn’t matter such as in mature, well defined assembly line processes where the vast majority of automation activities are focused.
People should stop listening to the economists and social scientists who pontificate about the looming automation crisis, and start listening to the engineers who actually have experience automating things. It is NOT an easy problem, and human brains are still incredibly valuable things to have at all levels of production. Thus it will remain until we build robots than can think generally and creatively. Which may never happen.
In the mantime, the only reason we automate something is if it is profitable to do so. And when we increase overall wealth through automation, it just creates more jobs elsewhere. In the same way, the mechanization of farming displaced millions of workers from farms, but the resulting economic growth created more jobs than were lost. I see no reason to assume that pattern won’t continue.
100% agree, well said. We’ve even gone back to rethinking some operations to re-introduce humans for some of our low volume multiple product type areas, due to the flexibility allowed by human labor. I’m sure there will be improvements in factory automation in the near future, but a degree or experience in computer integration and electronics will still be highly sought after for the foreseeable future at the factory level.
The drive towards factory automation is actually primarily motivated by two things: Improved operational efficiency through accuracy, repeatability and measurability, and the need to replacie workers as they retire. Specificsally, the drive towards digitization has a lot to with trying to nail down processes and document them before the oeople who have all the domain knowledge retire.
This is necessary because factories are having a hard time attracting new workers. This is exactly the opposite of the claim that there is some demand for factory jobs that no longer exist. The average age of floor workers in factories has been climbing for years, and I believe is aroun 50 now. This terrifies factory managers. When old Gus, the go-to guy to figure out why the line is rejecting 5% of its product retires, and there was no apprentice below him to learn all he knows, the factory loses a bunch of knowledge that was very expensive to acquire. So they want old Gus to have his kmowledge documented and embedded into a ditiAtion program before he leaves.
Digitization is the first step before you can automate. You need to capture your processes, document them, and cnvert them into process plans that can be executed by automation. This is turning put to be exceedingly hard in some cases, as quite often the process in place only works because the humans executing it constantly use their own judgement to solve problems and deviate from the plan when the plan deviates from reality.
In other words, before you can even digitize a process you have to discover what it actually is, then figure out how to remove the need for human judgement from it. The complexity of these processes can result in the failure of the digitization process. In that case, the factory has to essentially rebuild its processes around automation from the botttom up, and sometimes they discover that there are severe limits to how much you can automate.
Elon Musk thought he could build a fully automated car plant. He was sure of it, and he bet billions on it. Years of failures later, Musk now admits that he grossly underestimated the need for humans in the production process.