By not leveraging the machine you're placing yourself at a disadvantage, as you can only use the code others wrote and that will almost never be exactly what you need.
How much time is lost to custom code, learning to make custom code, and maintaining or replacing custom code the next time around? How much use is there in sharing standard processes and procedures, and how much time and, ultimately, cost is saved by outsourcing that to specialists?
I *can* program, to a limited degree; I still don't bother writing anything more substantial than a grep filter. C and C++ and the like are useless for modern purposes (huge, highly-polymorphic applications), and modern languages like C# actually eliminate half the effort of making your code fit together (reflection is fucking fantastic); and the effort I've put into learning C# and writing my own code... I am legitimately considering just spending half my salary to pay people to do this. I could have a prototype for the game I'm trying to write by now if I just did that--the assets actually cost more than the code.
All of these things are things I'd be able to do rapidly if I put in the up-front investment. The long-term returns haven't been substantial. My ability to use a simple shell script to break down a pile of text is substantial, and that took me 20 minutes to learn on my own by trial and error.
There's also the dufus claiming this is exactly why COPPA was enacted in the first place. COPPA was enacted to prevent web site operators from gathering tracking and identifying data about children; it wasn't enacted to prevent modern AI systems from collecting behavioral data to adjust themselves to the user's needs and preferences, which is exactly what these devices do. Their core functionality is "Oh, you're asking me for something specific that fits a sea of general things, but I know what you mean and will do the right thing!" They might not be great at it, but that's what they're for. COPPA is there to prevent collection and trade of data to marketers--i.e. for *their* use, not for *our* use.
Propylene glycol, you're right. I mixed up the wood preservative with the food-grade antifreeze preservative. Propylene glycol's toxicity is even lower, so my argument stands.
Man I've been having a bad day. That's twice now someone's had to correct me on something.
I didn't make that assumption; I made some examples of time scales. It is a good assumption that robots will improve at a certain pace and then slow down, and thus that getting in early and getting in late are both bad timing.
Tobacco is bad, sure, but who the fuck knows what solvents, preservatives, and fuck-knows-what goes into those little bottles that you breathe all... the... way... in
Polyethylene glycol, menthol, and a little nicotine. Essentially harmless stuff, unless you try chugging it. Nicotine's toxicity is rather high (it's a few thousand times the psychoactive dose, just the psychoactive dose is tiny); polyethylene glycol is pretty harmless (a soda can's worth per day would have negative health effects, mostly straining your kidneys; high doses are acutely toxic); and menthol is just mint oil.
That's funny because tesla cars are the least-dangerous in the most-severe collisions. Even when the battery is all fucked up, the fire is isolated away from the passenger compartment; and high-impact collisions transfer much less energy to the driver and passengers, thanks to enormous crumple zones.
I like math literacy, and I think we should fix the math curriculum. I'm largely focused on more effective methods of teaching arithmetic, algebra, and geometry, notably via Soroban and Anzan methods common in Japan. There is no reason we need to drill old, slow, and fanciful mathematics into kids's heads when we can drill old, *fast*, and efficient mathematics in a third of the time. That way they'd be able to rapidly perform basic, trivial computations and interact with the world in general (GEOMETRY, dude, how much cake mix if you're using a 9 inch round pan instead of an 11 x 14 rectangle?) by the time they're out of fourth grade, and we could all move on to something useful.
Don't get me wrong, Calc 2 is useful; so is integrated systems programming in C#.NET and model-view-controller interfaces using Bootstrap.JS. It's not useful to god damn *everyone*, whereas geometry and adding two numbers *is*; and the path to higher math is considerably shorter and better-founded when you've got strong arithmetic and algebraic skills. We do not need to graduate every high schooler with high-level multi-variable integration and a firm grasp of theoretical quantum physics.
Well. Don't leave your house. I personally like having my servants bag my groceries, stock the grocery shelves for me in the first place, prepare my hamburgers, and otherwise wait on me hand and foot. I'm not rightly sure what I would do if the grocery truck just dropped its trailer off at Giant and nobody bothered to unload the food, put it on the shelves, or sell it to me. I guess I could go to McDonalds, where I would stand in an empty building in which nobody has unloaded the food trucks, nobody will make me a burger, and nobody will allow me to order a burger to be made.
That kind of makes 24-hour service a lot cheaper, which is a *big* win because it means we reach a technical progress level at which we can have 24-hour service. This is similar to when we developed better steelmaking processes and got machines to dig ditches, and suddenly the taxes required to provide in-home running water dropped from 92% of your income to 5% of your income (it's like 1.2% of average income today--around $600/year here versus $54k median income). It's not identical: reducing the cost of steelmaking made steel viable for many things, thus creating a lot of steelmaking jobs (the computer revolution did this with a *lot* of information services); McDonalds automation is more on the Industrial Revolution model, which is *dangerous*.
That is 168 hours of operation or 4.2 employees at 40 hrs/week. So that robot can replace 4 employees at $16k/year and takes the ROI down to 6 months.
Technically, it replaces one *extremely* inefficient employee at the given time, and 4 full-time employees over the course of the year. What you said is thus correct, and I want to point out that we've essentially made those 4 employee's jobs *extremely* efficient, thus requiring less labor--fewer of them. This is because a single employee at a night-time McDonalds is underutilized: 6 employees with 160 sales per hour during the day is as efficient as 1 employee with 26 sales per hour at night; with the average sale volume being less, that night employee is inefficient.
We don't want companies maximizing pay - we want them minimizing cost. Cheaper shoes are good for everyone who buys shoes.
I concluded the same when I designed my own economic theory *before* bothering to glance at formal economics. (It turns out modern economics are about measuring, and not about the function of an economy; I can generate and explain a lot of modern theories using my own as a sort of economic fundamental theory.) Most people don't understand technical progress and how it affects an economy; and they don't understand demand-side economics.
I've shown people how demand-side economics works, and drawn up a hypothetical scenario where we blockade China and bring Manufacture back to America. The *very* *first* *thing* you'll notice is we have 5.6% UE4 unemployment, and so we don't have the sheer labor to make all this Chinese crap; but that's naive. When you start doing the wage calculations and looking at how this change affects consumer buying power, you realize we don't have the buying capacity to purchase modern health care, IT services (cell phones, spotify, etc.), and other luxuries of the day, and so the middle class and the poor just get shitloads poorer. In the end, bringing those jobs back to America weakens the consumer's buying power so much that tons of logistics, shipping, and retail jobs go away; somewhere between 15 million and 40 million American jobs are unsupportable under that model.
That means cheap Chinese wages have created tens of millions of American jobs.
Nobody thinks in that way. They think, oh, the businesses should pay more wages, and not charge me for it... where are they going to get the money to pay those wages? Trickle-down economics: you work hard, you get an education, you start a business, and you become rich; everyone forgets you do it by making sure somebody else's job fails.
But delaying investment only works in the absence of competition. 4 years to get your ROI now means that in year 5 you're down to maintenance costs for your robots.
At 25 cents per hour and a 14% labor proportion, that's only $1.78 per employee per hour more revenue required to keep employees versus getting machines. With 160 sales per hour on average (over 1,900 sales per day over the reasonable 12-hour daytime period) and $8/sale average, that's $1,280/hr income and $0.0111 per sale. With 6 employees, that's 7 cents per sale. You think McDonalds can't be competitive because they have to charge a penny more for a soda and 7 cents more for that supersized Extra Value Meal?
You're not "down to maintenance costs for your robots" in 5 years; you're dealing with an amortized TCO across the life of the machine. There's a spike in cost up-front, which you might be able to afford by digging into your bank accounts or by taking a loan; then you have to recover that loss with later profits.
If that's the year your competition decides to make their investment in robots, you can drop your price due to how little it now costs you to make burgers.
Only if you previously *raised* your prices to get ahead of the TCO amortization, thus cutting out your ability to compete.
Over the long run life of the machine, the business which gets in later and finds that they're saving $4 per employee per hour. They're trying to make labor represent 14% of their revenue, so that's $23.53/hr per employee that they can now reduce their prices. That means 5 years down the line, McDonalds starts rolling out price drops of $0.147 per employee per sale, instead of a penny--and with the 6-employee average staff referenced, that's 88 cents per sale.
So now Wendy's has jumped in in 2015 and reduced their $8 Single Biggie Meal to $7.93; McDonalds waited until 2020 and reduced their $8 Quarter Pounder Large Extra Value Meal to $7.12. Wendy's can't switch to the newer, more-efficient machines until 2028 or they won't get the ROI--they'll take a big loss on the whole deal--and even McDonalds is guessing the savings might be $5 per employee by then compared to 2015 (i.e. they think the cost of machines is going to level off around 2020).
That's how it plays out in the real world. Businesses do this all the time. Early-adopters are the ones who usually go bankrupt first.
You're only strengthening my argument, you know. I have a habit of ignoring things which strengthen my argument but which require an unsubstantiated handwave. "Benefits and taxes are really expensive, too!" The only thing I can quantify there is the $1,056 of OASDI and HI paid to social security out of payroll taxes on top of the employee's paycheck; I don't know how many part-time employees McDonalds has (any assembly of 40 hours is one full-time job, and people with fewer hours are underemployed and count as *less* than one job), so I don't know how many of these $16,500 jobs also incur a benefits cost and could only make something up.
Adidas's GROSS profits are increasing. "Gross Profit" is total revenue--sales. Outside of the financial industry, "Gross Profit" is a weasel-word used to mislead people: we usually think of "profit" as "net profit," which is the gross profit minus operating expenses. This makes sense because operating expenses include wages (employee gross profits--net is minus taxes), supply line (other business's gross profits), and outsourced business services (again, other business's gross profits). If you put all business net profits together with all employee gross income, you get the total income.
Gross profits increase if your employee wages get more expensive and you thus adjust the price of your product. Your net profit can actually decrease under this situation, leaving your business with less money at the end of the year.
As for this:
I would love to know why Adidas can't afford to pay decent wages?
Businesses don't pay wages. Consumers pay wages.
Wages in aggregate across the entire production process are the base cost: at the end of the day, the product will absolutely sell for no less expensive than that. Volume deals push the price closer to the cost, such as when GM tries to bid for 100,000,000 tonnes of steel per year, and the steel mills contract with the coal and ore companies contingent on winning the GM contract, and everyone slims their profit margin because taking $1 per tonne on 100,000,000 tonnes is still $100,000,000 versus trying to profit $20 per tonne and selling 0 tonnes to GM. No matter how hard you compact that down--get it down to tenths of a cent per unit and 0.1% profit margins--you'll get no lower than the wage-labor costs of all employees involved in the entire supply chain.
Raise the wage-labor cost such that the steel costs $20 more per tonne and the price the steel mill will need to charge goes up by $20. With GM making passenger cars weighing 1.5 tonnes in steel, those cars cost $30 more. Either GM absorbs the cost in the form of lower profit margins (in which case, GM, as the consumer of steel, pays the wage of the steel mill) or GM holds its profit margin (usually 7%-13%; was -7.5% in 2013--they took a loss, which the big profit margins help protect against) and the end consumer pays for the wage raise.
In the case of Adidas, when shoe-maker labor increases, they can either raise prices or lower profits. Adidas's profits barely offset their loss years, with a five-year average of 4% and a five-year low of -7.5%. That means a 4% increase in labor costs--29 cents in minimum-wage increase in the United States, or a 14 cents increase on $3.50/hr Chinese labor--can put Adidas into permanent loss, ending in bankruptcy.
I could claim teaching everyone agricultural management is a good idea, and I would be wrong; of course a huge flock of neo-conservative anarchocapitalists would get behind me on that one, citing that we should all be able to independently make our own food, so a mandatory master's in farming is a good thing.
People don't need computer science education; they need education in operating a computer, and, as much as you want it to be true, programming is *not* operating, in the same way that *engineering a car* is not *driving*.
I think their movement to machines is actually a cost-savings measure, plus they're getting money from the Government to do so. The Government wants more industrial machines to move humans away from industrial accidents.
Labor costs only account for about 20% of the cost of a typical fast food item.
It's 14% of revenue. Industry standard. Below 14%, they start calling people in if business is expected to pick up; above 14%, they start sending people on break or dismissing them early. 20% would be huge.
People have done the math and McD's would only need to raise the price of a bag of french fries about 4cents to cover a $15 minimum wage.
14% of an $8 value meal is $1.12; $15 is 181% of $8.25. It's 92 cents extra for that extra value meal.
Even at 4 cents, that's 1 job lost per 412,500 boxes of small fries sold.
In reality, the total U.S. fast food industry revenue (total sales) in 2013 was $191 billion. That's $191,000,000,000. The cost of direct wages was 24.6%, which accounts for managers, district managers, corporate operations, and so forth; at the line, it's 14%, which *doesn't* account for minimum-wage workers (or anyone making $15/hr) up the chain.
Let's assume all minimum wage workers in the United States make $8.25/hr (MD standard, higher than $7.25/hr Federal standard) and that anyone making more than minimum wage makes more than $15/hr. THIS WILL SKEW MY NUMBERS LOW. Let's also ignore any non-direct wage, which will also skew my numbers low. That means *only* the cashier and burger flippers are getting raises; the managers, truck drivers, packing industry workers, and so forth aren't getting raises. If you don't work at the bottom rung in a Wendy's, fuck you.
14% of $191 billion is $26.74 billion. Multiply that by 1.81 and you get $48.40 billion. That's a difference of $21.66 billion dollars, or 1.3 million U.S. jobs lost. To put this into perspective: the United States gained 9 million jobs between 2010 and 2016; 1.5 years of population and job growth gone in one sweep is pretty big.
You have this ideal where it's like 53 cents extra for that $8 value meal, and you can afford 53 cents, so this isn't a big deal. Aggregate that 53 cents across the entire consumer base, and it turns into entire years's worth of wages for millions of people--wages that are going to fewer hands. It's still the same amount of spending per year, too: these people were getting paid before, they're getting paid more now, and others aren't getting paid at all; they don't magically have more consumer money to spend and create new jobs because they can't spend it until they earn it, at which point everything is slightly more expensive and the same impact occurs.
Remember: that 1.3 million jobs is *just* the direct impact of low-level line crew at McDonalds getting bumped from $8.25 to $15. It ignores any manager salary raises; it ignores janitors at corporate; and it ignores any sub-$15 wages in the domestic portion of the supply chain (we are obviously not raising minimum wage for the Chinese). Any of these things will eliminate more jobs.
Sometimes I really wonder how these companies survive.
The answer is simple: they're run by executives who actually know what they're doing, and who aren't actually playing out the fanciful narrative you've described. Some businesses are projectized, and so don't have firm functional divisions; others operate in less-rigid ways, and so are able to restructure on-the-fly without the mass expense an old-style 1920s office would face when trying to reorganize. We've done a lot of research since 1970 on how to get things done and how to minimize the risk of change, both by avoiding it and by expending ~1% of the cost of potential change so that such changes end up costing ~5% as much if they do happen.
I get shuffled about, but it's a zero-cost operation. Somebody clicks a button and the manager responsible for my time is some other department head. When budget time comes, that guy's department pays the cost of my salary, and the balance sheets show different expenses from different areas. There is so much slack in the day that this operation doesn't even impede the normal work of whomever is assigned to it--we've come a long way from factory work, although not so far if you're a programmer (programmers spend nearly 100% of their time thinking about how to accomplish a task or accomplishing that task, and have much less slack time than other office workers).
Of course they'd want it managed as competently as possible. They have to bid low against competition, and every unpredicted and uncontrolled risk cuts into their profits on that. Failure to deliver on-budget bans you from government contracts with the agency for several years. Besides that, the agency awards a fixed-fee-plus-awards or time-plus-awards contract for this sort of work, meaning finishing at higher quality and less time generates a higher per-hour billable than stretching the work, and thus you can move on to the next project at higher per-hour billables if you can keep being awesome.
I actually wouldn't mind being PM on a modernization job like that. It's the sort of high-complexity, high-risk program that got me into project management in the first place: so many things can go wrong, from bad requirements gathering to bad delivery, and predicting and controlling all those risks is an amazing challenge.
Yup. This isn't really a valid argument against increasing the minimum wage.
At worst, it merely hastens the inevitable by a few years, but this is going to happen.
I addressed this. You repeat a line that comes from the thinking that jobs go away forever and no new jobs come. Circa 1790, 90% of American workers were farmers; today that's 2%, and a total of 11% of the workforce (including the farmers themselves) provides all the supporting infrastructure (energy, machines, pesticides, fertilizer, shipping, retail, marketing) to supply our food.
It's not "Hastening"; it's "Compacting." You're creating a situation where people become unemployed at a higher rate--more jobs lost per month--and replacement jobs come at a lower rate. Instead of shaking a little as we push up to 6% unemployment and then come back down to 5% over 5-10 years, we spike up to 30% unemployment over 2-3 years and then require some 70 years to recover--if our economy doesn't fucking collapse first.
You will incur enough injuries and blood loss in your life that, were I to take all that blood from your body today, you would die immediately.
If you do it slowly enough, the fast food prices won't rise as quickly as inflation, and consumer buying power will increase. That's what actually happens. To buy more stuff, we need more workers--operating the machines, of course--which means new jobs. We spend a smaller proportion of our income on the stuff we buy now, and the remainder goes to new things--since the 80s, we've moved our money from food and clothing onto more and better healthcare, as well as smart phones and electronic entertainment; and houses have gotten bigger, while cars have gained luxury, performance, and safety features while remaining roughly 56% of the median income.
If you do it quickly, you get an unemployment spike, which damages the economy. The bigger the spike, the longer it takes to recover, and the poorer your society comes out of it.
If you do it poorly, you get rough destruction of wealth. Raising minimum wage already concentrates wealth into a poor elite--some minimum-wage workers get richer, all other consumers become poorer, and we lose jobs: The cost of a burger increasing by 17 cents, with 282 billion burgers sold per year, is $50 million; that's over 3,000 $8.25/hr jobs. Raising minimum wage such that the old wage was cheaper than a machine *and* the new wage is more expensive ($8.25 wage becomes $15 wage; machine is $9.50) eliminates the minimum wage jobs and exchanges in more-expensive machines, so your economy takes it both ways.
A lot of people can't grok this because it's a continuous-operation function. Basically, people reason, "Hey, but the minimum-wage worker has more money to spend, and so you wouldn't lose any jobs!" By such reasoning, you have infinite money, and thus infinite jobs, and we are all fabulously wealthy (we are, but that's not the point). You have so much income *per* *time*, and the cost of purchasing certain goods increases, and so the number of goods increased *per* *time* decreases, thus the jobs decrease. Again: doubling down on this kind of damage by making wage workers non-competitive with machines is bad.
It gets even worse: normally, product price increases occur slightly more slowly than inflation for products whose costs have decreased. That is to say: If you displace 10% of the labor cost of making a hamburger, that hamburger will approach 1.8 times the price after 100% inflation--10% of its price doesn't keep up. As this money returns to the consumer, the consumer base becomes capable of paying the wage of another worker, and thus can buy new products. If you've pushed up the cost of labor, then it takes *longer* for those two things to intersect, and so the transitional period of unemployment extends: jobs lost to technical progress take more time to become new jobs.
So you're de-employing workers *quickly* (unemployment coming more rapidly); you're increasing the cost of goods instead of decreasing it (setting the far point of technical progress growth years farther out); and you're making human labor more expensive (requiring much more purchasing power movement back to the consumer's hands before replacement jobs are created--and reducing the total replacement jobs possible).
That's a recipe for an economic disaster and a permanent feedback loop to make a society poorer. It's one of the reasons I push for a Citizen's Dividend that migrates costs off wage-labor (reduce payroll taxes and replace minimum wage raises with a non-wage income basis): that plan increases the number of consumer take-home dollars per employer wage-labor dollars paid to have an employee. You can describe that as "decreasing costs" or "increasing consumer buying power"; if you stare long enough, you realize the two things are eventually the same.
Technical progress is what makes the middle-class, the poor, *and* the rich richer. It's what's given us the ability to *afford* modern healthcare, high-speed internet, wireless ph
They're going to replace employees with robots anyhow, I don't buy that increasing the minimum wage to whatever has anything to do with it.
Businesses have risk appetite and risk tolerance. Risk appetite is how much money they want to throw in a hole for a likely conversion to more money; risk tolerance is the point at which they will not throw more money into the hole because the return--whether or not it's coming--is sinking the ship too hard, and they're no longer interested in trying to squeeze out more promised droplets of gold.
Because of risk appetite, different businesses will implement labor-reducing changes at different times. Sure, you have an $8.25/hr employee now, and the machine costs $8/hr; but next year the machine should compare to a $7.25/hr employee, and in three years it should compare to a $5.50/hr employee. It seems to me that, over the ten-year period, you will come out with a higher profit if you wait three years before deploying expensive machines. These are $35,000 machines replacing $16,500 employees, so you need a little over 2 years to get a break-even ROI (replace benefits with maintenance).
To some businesses, switching onto machines right away seems like a good idea. Poor foresight I guess. Other businesses will vary between how they roll out--how long to delay, how fast to carry out the roll-out, etc. That means moving everyone out of their jobs and getting machines in here could take a decade or more if wages are competitive with machines and we believe machines will get cheaper. The risk of moving onto machines isn't offset by the 25 cents savings, and the potential return for paying that 25 cents for the next few years is that you turn it into a 4 dollar savings instead.
This breaks when you suddenly make labor expensive.
Now instead of $8.25/hr vs $8/hr, you're doing $15/hr vs $8/hr. In one year, avoiding the 25 cent savings means $500 per employee per year; but at $15/hr, you're losing $7,000 per employee per year for not going in right now. That's going to hit risk tolerances a lot faster, and jobs are going away much more rapidly in those conditions.
It's even worse if machines are *more* expensive than people: you get the price increase that comes with, say, $11/hr (machine) labor, but you fire a bunch of $8.25/hr human labor. Normally, we replace a high-labor process with a lower-labor one and make cost savings, leading to a reduction of prices, leaving more money in consumer pockets, allowing more purchasing, creating new jobs to make the new stuff we're buying. If the machines are more expensive than wage-workers before the wage bump, then costs go *up*, and consumer ability to buy goes *down*: rather than reacting to the reduction of jobs by creating new jobs, the consumer base reacts to the increase in cost by not being able to financially support the wages of *even* *more* *jobs*.
In 1790, 90% of Americans laborers (in a ~58% labor force) were farmers; we've replaced most of them with machines, and they now make up 2% of the labor force, and about 11% of consumer spending in total goes toward food to cover those farmers, the people building and maintaining farm equipment, logistics and sales moving that kind of thing, chemical companies making fertilizers and pesticides, and oil mining and refining to get the fuel for energy to drive all this. That means 18% of the labor involved in making food is on the farm, and 82% is in supporting infrastructure. You'll notice we don't have an 82% unemployment rate today; and automated fast food won't destroy our job market unless the method by which we transition is damaging--which this particular method *is*.
It doesn't fail as a fitness device. They acknowledge that spot-measurement of heart rate may be inaccurate; and also cite that such inaccuracies--even severe ones--don't affect its ability to use HRM data to calculate calories burned because it uses that data to measure statistical trends. In other words: FitBit isn't measuring (b) BPM and correlating that to (c) calories; it's measuring increases, decreases, and stable levels in heart rate, and determining if you're exerting yourself or not, and by how much (how rapidly and stably is your heart rate rising or falling?).
I've used FitBit and the FitBit Aria scale (which is, seriously, inaccurate with its body fat percentage measurement) along with calorie counting, and noticed a long-term trend of weight loss if I burned more calories than I took in. Weight fluctuates such that I might weigh 2 pounds more one day than the next, although I've held it pretty stable (half a pound fluctuation) by measuring at the same time under the same conditions (notably, in the morning, before eating, before showering, after pissing--minimal hydration level and post-fast); however, I lost 20 pounds in 3 months with moderate activity (walking for 20 minutes per day after lunch). I ate more when I burned more calories, and made sure I was coming under most of the time. As long as I kept that trend--even when I was burning 2700kcal/day and eating 2500, notably Popeye's Fried Chicken 1,100kcal lunch 3 days per week--my gut fat went down and my weight decreased.
Based on that, I would say they got the general trend correct, surprisingly so.
Slashdot Mirror
By not leveraging the machine you're placing yourself at a disadvantage, as you can only use the code others wrote and that will almost never be exactly what you need.
How much time is lost to custom code, learning to make custom code, and maintaining or replacing custom code the next time around? How much use is there in sharing standard processes and procedures, and how much time and, ultimately, cost is saved by outsourcing that to specialists?
I *can* program, to a limited degree; I still don't bother writing anything more substantial than a grep filter. C and C++ and the like are useless for modern purposes (huge, highly-polymorphic applications), and modern languages like C# actually eliminate half the effort of making your code fit together (reflection is fucking fantastic); and the effort I've put into learning C# and writing my own code... I am legitimately considering just spending half my salary to pay people to do this. I could have a prototype for the game I'm trying to write by now if I just did that--the assets actually cost more than the code.
All of these things are things I'd be able to do rapidly if I put in the up-front investment. The long-term returns haven't been substantial. My ability to use a simple shell script to break down a pile of text is substantial, and that took me 20 minutes to learn on my own by trial and error.
There's also the dufus claiming this is exactly why COPPA was enacted in the first place. COPPA was enacted to prevent web site operators from gathering tracking and identifying data about children; it wasn't enacted to prevent modern AI systems from collecting behavioral data to adjust themselves to the user's needs and preferences, which is exactly what these devices do. Their core functionality is "Oh, you're asking me for something specific that fits a sea of general things, but I know what you mean and will do the right thing!" They might not be great at it, but that's what they're for. COPPA is there to prevent collection and trade of data to marketers--i.e. for *their* use, not for *our* use.
Propylene glycol, you're right. I mixed up the wood preservative with the food-grade antifreeze preservative. Propylene glycol's toxicity is even lower, so my argument stands.
Man I've been having a bad day. That's twice now someone's had to correct me on something.
I didn't make that assumption; I made some examples of time scales. It is a good assumption that robots will improve at a certain pace and then slow down, and thus that getting in early and getting in late are both bad timing.
Tobacco is bad, sure, but who the fuck knows what solvents, preservatives, and fuck-knows-what goes into those little bottles that you breathe all... the... way... in
Polyethylene glycol, menthol, and a little nicotine. Essentially harmless stuff, unless you try chugging it. Nicotine's toxicity is rather high (it's a few thousand times the psychoactive dose, just the psychoactive dose is tiny); polyethylene glycol is pretty harmless (a soda can's worth per day would have negative health effects, mostly straining your kidneys; high doses are acutely toxic); and menthol is just mint oil.
That's funny because tesla cars are the least-dangerous in the most-severe collisions. Even when the battery is all fucked up, the fire is isolated away from the passenger compartment; and high-impact collisions transfer much less energy to the driver and passengers, thanks to enormous crumple zones.
You're right. My business knowledge is slightly-less-complete than my economics knowledge. The rest still stands.
I like math literacy, and I think we should fix the math curriculum. I'm largely focused on more effective methods of teaching arithmetic, algebra, and geometry, notably via Soroban and Anzan methods common in Japan. There is no reason we need to drill old, slow, and fanciful mathematics into kids's heads when we can drill old, *fast*, and efficient mathematics in a third of the time. That way they'd be able to rapidly perform basic, trivial computations and interact with the world in general (GEOMETRY, dude, how much cake mix if you're using a 9 inch round pan instead of an 11 x 14 rectangle?) by the time they're out of fourth grade, and we could all move on to something useful.
Don't get me wrong, Calc 2 is useful; so is integrated systems programming in C#.NET and model-view-controller interfaces using Bootstrap.JS. It's not useful to god damn *everyone*, whereas geometry and adding two numbers *is*; and the path to higher math is considerably shorter and better-founded when you've got strong arithmetic and algebraic skills. We do not need to graduate every high schooler with high-level multi-variable integration and a firm grasp of theoretical quantum physics.
Well. Don't leave your house. I personally like having my servants bag my groceries, stock the grocery shelves for me in the first place, prepare my hamburgers, and otherwise wait on me hand and foot. I'm not rightly sure what I would do if the grocery truck just dropped its trailer off at Giant and nobody bothered to unload the food, put it on the shelves, or sell it to me. I guess I could go to McDonalds, where I would stand in an empty building in which nobody has unloaded the food trucks, nobody will make me a burger, and nobody will allow me to order a burger to be made.
I like eating.
That kind of makes 24-hour service a lot cheaper, which is a *big* win because it means we reach a technical progress level at which we can have 24-hour service. This is similar to when we developed better steelmaking processes and got machines to dig ditches, and suddenly the taxes required to provide in-home running water dropped from 92% of your income to 5% of your income (it's like 1.2% of average income today--around $600/year here versus $54k median income). It's not identical: reducing the cost of steelmaking made steel viable for many things, thus creating a lot of steelmaking jobs (the computer revolution did this with a *lot* of information services); McDonalds automation is more on the Industrial Revolution model, which is *dangerous*.
That is 168 hours of operation or 4.2 employees at 40 hrs/week. So that robot can replace 4 employees at $16k/year and takes the ROI down to 6 months.
Technically, it replaces one *extremely* inefficient employee at the given time, and 4 full-time employees over the course of the year. What you said is thus correct, and I want to point out that we've essentially made those 4 employee's jobs *extremely* efficient, thus requiring less labor--fewer of them. This is because a single employee at a night-time McDonalds is underutilized: 6 employees with 160 sales per hour during the day is as efficient as 1 employee with 26 sales per hour at night; with the average sale volume being less, that night employee is inefficient.
We don't want companies maximizing pay - we want them minimizing cost. Cheaper shoes are good for everyone who buys shoes.
I concluded the same when I designed my own economic theory *before* bothering to glance at formal economics. (It turns out modern economics are about measuring, and not about the function of an economy; I can generate and explain a lot of modern theories using my own as a sort of economic fundamental theory.) Most people don't understand technical progress and how it affects an economy; and they don't understand demand-side economics.
I've shown people how demand-side economics works, and drawn up a hypothetical scenario where we blockade China and bring Manufacture back to America. The *very* *first* *thing* you'll notice is we have 5.6% UE4 unemployment, and so we don't have the sheer labor to make all this Chinese crap; but that's naive. When you start doing the wage calculations and looking at how this change affects consumer buying power, you realize we don't have the buying capacity to purchase modern health care, IT services (cell phones, spotify, etc.), and other luxuries of the day, and so the middle class and the poor just get shitloads poorer. In the end, bringing those jobs back to America weakens the consumer's buying power so much that tons of logistics, shipping, and retail jobs go away; somewhere between 15 million and 40 million American jobs are unsupportable under that model.
That means cheap Chinese wages have created tens of millions of American jobs.
Nobody thinks in that way. They think, oh, the businesses should pay more wages, and not charge me for it... where are they going to get the money to pay those wages? Trickle-down economics: you work hard, you get an education, you start a business, and you become rich; everyone forgets you do it by making sure somebody else's job fails.
But delaying investment only works in the absence of competition. 4 years to get your ROI now means that in year 5 you're down to maintenance costs for your robots.
At 25 cents per hour and a 14% labor proportion, that's only $1.78 per employee per hour more revenue required to keep employees versus getting machines. With 160 sales per hour on average (over 1,900 sales per day over the reasonable 12-hour daytime period) and $8/sale average, that's $1,280/hr income and $0.0111 per sale. With 6 employees, that's 7 cents per sale. You think McDonalds can't be competitive because they have to charge a penny more for a soda and 7 cents more for that supersized Extra Value Meal?
You're not "down to maintenance costs for your robots" in 5 years; you're dealing with an amortized TCO across the life of the machine. There's a spike in cost up-front, which you might be able to afford by digging into your bank accounts or by taking a loan; then you have to recover that loss with later profits.
If that's the year your competition decides to make their investment in robots, you can drop your price due to how little it now costs you to make burgers.
Only if you previously *raised* your prices to get ahead of the TCO amortization, thus cutting out your ability to compete.
Over the long run life of the machine, the business which gets in later and finds that they're saving $4 per employee per hour. They're trying to make labor represent 14% of their revenue, so that's $23.53/hr per employee that they can now reduce their prices. That means 5 years down the line, McDonalds starts rolling out price drops of $0.147 per employee per sale, instead of a penny--and with the 6-employee average staff referenced, that's 88 cents per sale.
So now Wendy's has jumped in in 2015 and reduced their $8 Single Biggie Meal to $7.93; McDonalds waited until 2020 and reduced their $8 Quarter Pounder Large Extra Value Meal to $7.12. Wendy's can't switch to the newer, more-efficient machines until 2028 or they won't get the ROI--they'll take a big loss on the whole deal--and even McDonalds is guessing the savings might be $5 per employee by then compared to 2015 (i.e. they think the cost of machines is going to level off around 2020).
That's how it plays out in the real world. Businesses do this all the time. Early-adopters are the ones who usually go bankrupt first.
You're only strengthening my argument, you know. I have a habit of ignoring things which strengthen my argument but which require an unsubstantiated handwave. "Benefits and taxes are really expensive, too!" The only thing I can quantify there is the $1,056 of OASDI and HI paid to social security out of payroll taxes on top of the employee's paycheck; I don't know how many part-time employees McDonalds has (any assembly of 40 hours is one full-time job, and people with fewer hours are underemployed and count as *less* than one job), so I don't know how many of these $16,500 jobs also incur a benefits cost and could only make something up.
Adidas's GROSS profits are increasing. "Gross Profit" is total revenue--sales. Outside of the financial industry, "Gross Profit" is a weasel-word used to mislead people: we usually think of "profit" as "net profit," which is the gross profit minus operating expenses. This makes sense because operating expenses include wages (employee gross profits--net is minus taxes), supply line (other business's gross profits), and outsourced business services (again, other business's gross profits). If you put all business net profits together with all employee gross income, you get the total income.
Gross profits increase if your employee wages get more expensive and you thus adjust the price of your product. Your net profit can actually decrease under this situation, leaving your business with less money at the end of the year.
As for this:
I would love to know why Adidas can't afford to pay decent wages?
Businesses don't pay wages. Consumers pay wages.
Wages in aggregate across the entire production process are the base cost: at the end of the day, the product will absolutely sell for no less expensive than that. Volume deals push the price closer to the cost, such as when GM tries to bid for 100,000,000 tonnes of steel per year, and the steel mills contract with the coal and ore companies contingent on winning the GM contract, and everyone slims their profit margin because taking $1 per tonne on 100,000,000 tonnes is still $100,000,000 versus trying to profit $20 per tonne and selling 0 tonnes to GM. No matter how hard you compact that down--get it down to tenths of a cent per unit and 0.1% profit margins--you'll get no lower than the wage-labor costs of all employees involved in the entire supply chain.
Raise the wage-labor cost such that the steel costs $20 more per tonne and the price the steel mill will need to charge goes up by $20. With GM making passenger cars weighing 1.5 tonnes in steel, those cars cost $30 more. Either GM absorbs the cost in the form of lower profit margins (in which case, GM, as the consumer of steel, pays the wage of the steel mill) or GM holds its profit margin (usually 7%-13%; was -7.5% in 2013--they took a loss, which the big profit margins help protect against) and the end consumer pays for the wage raise.
In the case of Adidas, when shoe-maker labor increases, they can either raise prices or lower profits. Adidas's profits barely offset their loss years, with a five-year average of 4% and a five-year low of -7.5%. That means a 4% increase in labor costs--29 cents in minimum-wage increase in the United States, or a 14 cents increase on $3.50/hr Chinese labor--can put Adidas into permanent loss, ending in bankruptcy.
How about we get out of this stupid fascination with our favorite pet topic.
While calling all policy makers and education leaders to consider 'computer science education for all' is a good thing
Begging the question: why is it a good thing? Half of the current CompSci grads don't even get CompSci jobs.
I could claim teaching everyone agricultural management is a good idea, and I would be wrong; of course a huge flock of neo-conservative anarchocapitalists would get behind me on that one, citing that we should all be able to independently make our own food, so a mandatory master's in farming is a good thing.
People don't need computer science education; they need education in operating a computer, and, as much as you want it to be true, programming is *not* operating, in the same way that *engineering a car* is not *driving*.
I think their movement to machines is actually a cost-savings measure, plus they're getting money from the Government to do so. The Government wants more industrial machines to move humans away from industrial accidents.
Bullshit. This time is Industrial Revolution model, and I already know how to control that model.
Labor costs only account for about 20% of the cost of a typical fast food item.
It's 14% of revenue. Industry standard. Below 14%, they start calling people in if business is expected to pick up; above 14%, they start sending people on break or dismissing them early. 20% would be huge.
People have done the math and McD's would only need to raise the price of a bag of french fries about 4cents to cover a $15 minimum wage.
14% of an $8 value meal is $1.12; $15 is 181% of $8.25. It's 92 cents extra for that extra value meal.
Even at 4 cents, that's 1 job lost per 412,500 boxes of small fries sold.
In reality, the total U.S. fast food industry revenue (total sales) in 2013 was $191 billion. That's $191,000,000,000. The cost of direct wages was 24.6%, which accounts for managers, district managers, corporate operations, and so forth; at the line, it's 14%, which *doesn't* account for minimum-wage workers (or anyone making $15/hr) up the chain.
Let's assume all minimum wage workers in the United States make $8.25/hr (MD standard, higher than $7.25/hr Federal standard) and that anyone making more than minimum wage makes more than $15/hr. THIS WILL SKEW MY NUMBERS LOW. Let's also ignore any non-direct wage, which will also skew my numbers low. That means *only* the cashier and burger flippers are getting raises; the managers, truck drivers, packing industry workers, and so forth aren't getting raises. If you don't work at the bottom rung in a Wendy's, fuck you.
14% of $191 billion is $26.74 billion. Multiply that by 1.81 and you get $48.40 billion. That's a difference of $21.66 billion dollars, or 1.3 million U.S. jobs lost. To put this into perspective: the United States gained 9 million jobs between 2010 and 2016; 1.5 years of population and job growth gone in one sweep is pretty big.
You have this ideal where it's like 53 cents extra for that $8 value meal, and you can afford 53 cents, so this isn't a big deal. Aggregate that 53 cents across the entire consumer base, and it turns into entire years's worth of wages for millions of people--wages that are going to fewer hands. It's still the same amount of spending per year, too: these people were getting paid before, they're getting paid more now, and others aren't getting paid at all; they don't magically have more consumer money to spend and create new jobs because they can't spend it until they earn it, at which point everything is slightly more expensive and the same impact occurs.
Remember: that 1.3 million jobs is *just* the direct impact of low-level line crew at McDonalds getting bumped from $8.25 to $15. It ignores any manager salary raises; it ignores janitors at corporate; and it ignores any sub-$15 wages in the domestic portion of the supply chain (we are obviously not raising minimum wage for the Chinese). Any of these things will eliminate more jobs.
Sometimes I really wonder how these companies survive.
The answer is simple: they're run by executives who actually know what they're doing, and who aren't actually playing out the fanciful narrative you've described. Some businesses are projectized, and so don't have firm functional divisions; others operate in less-rigid ways, and so are able to restructure on-the-fly without the mass expense an old-style 1920s office would face when trying to reorganize. We've done a lot of research since 1970 on how to get things done and how to minimize the risk of change, both by avoiding it and by expending ~1% of the cost of potential change so that such changes end up costing ~5% as much if they do happen.
I get shuffled about, but it's a zero-cost operation. Somebody clicks a button and the manager responsible for my time is some other department head. When budget time comes, that guy's department pays the cost of my salary, and the balance sheets show different expenses from different areas. There is so much slack in the day that this operation doesn't even impede the normal work of whomever is assigned to it--we've come a long way from factory work, although not so far if you're a programmer (programmers spend nearly 100% of their time thinking about how to accomplish a task or accomplishing that task, and have much less slack time than other office workers).
Of course they'd want it managed as competently as possible. They have to bid low against competition, and every unpredicted and uncontrolled risk cuts into their profits on that. Failure to deliver on-budget bans you from government contracts with the agency for several years. Besides that, the agency awards a fixed-fee-plus-awards or time-plus-awards contract for this sort of work, meaning finishing at higher quality and less time generates a higher per-hour billable than stretching the work, and thus you can move on to the next project at higher per-hour billables if you can keep being awesome.
Yeah, we should rewrite everything!
I actually wouldn't mind being PM on a modernization job like that. It's the sort of high-complexity, high-risk program that got me into project management in the first place: so many things can go wrong, from bad requirements gathering to bad delivery, and predicting and controlling all those risks is an amazing challenge.
Yup. This isn't really a valid argument against increasing the minimum wage.
At worst, it merely hastens the inevitable by a few years, but this is going to happen.
I addressed this. You repeat a line that comes from the thinking that jobs go away forever and no new jobs come. Circa 1790, 90% of American workers were farmers; today that's 2%, and a total of 11% of the workforce (including the farmers themselves) provides all the supporting infrastructure (energy, machines, pesticides, fertilizer, shipping, retail, marketing) to supply our food.
It's not "Hastening"; it's "Compacting." You're creating a situation where people become unemployed at a higher rate--more jobs lost per month--and replacement jobs come at a lower rate. Instead of shaking a little as we push up to 6% unemployment and then come back down to 5% over 5-10 years, we spike up to 30% unemployment over 2-3 years and then require some 70 years to recover--if our economy doesn't fucking collapse first.
You will incur enough injuries and blood loss in your life that, were I to take all that blood from your body today, you would die immediately.
If you do it slowly enough, the fast food prices won't rise as quickly as inflation, and consumer buying power will increase. That's what actually happens. To buy more stuff, we need more workers--operating the machines, of course--which means new jobs. We spend a smaller proportion of our income on the stuff we buy now, and the remainder goes to new things--since the 80s, we've moved our money from food and clothing onto more and better healthcare, as well as smart phones and electronic entertainment; and houses have gotten bigger, while cars have gained luxury, performance, and safety features while remaining roughly 56% of the median income.
If you do it quickly, you get an unemployment spike, which damages the economy. The bigger the spike, the longer it takes to recover, and the poorer your society comes out of it.
If you do it poorly, you get rough destruction of wealth. Raising minimum wage already concentrates wealth into a poor elite--some minimum-wage workers get richer, all other consumers become poorer, and we lose jobs: The cost of a burger increasing by 17 cents, with 282 billion burgers sold per year, is $50 million; that's over 3,000 $8.25/hr jobs. Raising minimum wage such that the old wage was cheaper than a machine *and* the new wage is more expensive ($8.25 wage becomes $15 wage; machine is $9.50) eliminates the minimum wage jobs and exchanges in more-expensive machines, so your economy takes it both ways.
A lot of people can't grok this because it's a continuous-operation function. Basically, people reason, "Hey, but the minimum-wage worker has more money to spend, and so you wouldn't lose any jobs!" By such reasoning, you have infinite money, and thus infinite jobs, and we are all fabulously wealthy (we are, but that's not the point). You have so much income *per* *time*, and the cost of purchasing certain goods increases, and so the number of goods increased *per* *time* decreases, thus the jobs decrease. Again: doubling down on this kind of damage by making wage workers non-competitive with machines is bad.
It gets even worse: normally, product price increases occur slightly more slowly than inflation for products whose costs have decreased. That is to say: If you displace 10% of the labor cost of making a hamburger, that hamburger will approach 1.8 times the price after 100% inflation--10% of its price doesn't keep up. As this money returns to the consumer, the consumer base becomes capable of paying the wage of another worker, and thus can buy new products. If you've pushed up the cost of labor, then it takes *longer* for those two things to intersect, and so the transitional period of unemployment extends: jobs lost to technical progress take more time to become new jobs.
So you're de-employing workers *quickly* (unemployment coming more rapidly); you're increasing the cost of goods instead of decreasing it (setting the far point of technical progress growth years farther out); and you're making human labor more expensive (requiring much more purchasing power movement back to the consumer's hands before replacement jobs are created--and reducing the total replacement jobs possible).
That's a recipe for an economic disaster and a permanent feedback loop to make a society poorer. It's one of the reasons I push for a Citizen's Dividend that migrates costs off wage-labor (reduce payroll taxes and replace minimum wage raises with a non-wage income basis): that plan increases the number of consumer take-home dollars per employer wage-labor dollars paid to have an employee. You can describe that as "decreasing costs" or "increasing consumer buying power"; if you stare long enough, you realize the two things are eventually the same.
Technical progress is what makes the middle-class, the poor, *and* the rich richer. It's what's given us the ability to *afford* modern healthcare, high-speed internet, wireless ph
They're going to replace employees with robots anyhow, I don't buy that increasing the minimum wage to whatever has anything to do with it.
Businesses have risk appetite and risk tolerance. Risk appetite is how much money they want to throw in a hole for a likely conversion to more money; risk tolerance is the point at which they will not throw more money into the hole because the return--whether or not it's coming--is sinking the ship too hard, and they're no longer interested in trying to squeeze out more promised droplets of gold.
Because of risk appetite, different businesses will implement labor-reducing changes at different times. Sure, you have an $8.25/hr employee now, and the machine costs $8/hr; but next year the machine should compare to a $7.25/hr employee, and in three years it should compare to a $5.50/hr employee. It seems to me that, over the ten-year period, you will come out with a higher profit if you wait three years before deploying expensive machines. These are $35,000 machines replacing $16,500 employees, so you need a little over 2 years to get a break-even ROI (replace benefits with maintenance).
To some businesses, switching onto machines right away seems like a good idea. Poor foresight I guess. Other businesses will vary between how they roll out--how long to delay, how fast to carry out the roll-out, etc. That means moving everyone out of their jobs and getting machines in here could take a decade or more if wages are competitive with machines and we believe machines will get cheaper. The risk of moving onto machines isn't offset by the 25 cents savings, and the potential return for paying that 25 cents for the next few years is that you turn it into a 4 dollar savings instead.
This breaks when you suddenly make labor expensive.
Now instead of $8.25/hr vs $8/hr, you're doing $15/hr vs $8/hr. In one year, avoiding the 25 cent savings means $500 per employee per year; but at $15/hr, you're losing $7,000 per employee per year for not going in right now. That's going to hit risk tolerances a lot faster, and jobs are going away much more rapidly in those conditions.
It's even worse if machines are *more* expensive than people: you get the price increase that comes with, say, $11/hr (machine) labor, but you fire a bunch of $8.25/hr human labor. Normally, we replace a high-labor process with a lower-labor one and make cost savings, leading to a reduction of prices, leaving more money in consumer pockets, allowing more purchasing, creating new jobs to make the new stuff we're buying. If the machines are more expensive than wage-workers before the wage bump, then costs go *up*, and consumer ability to buy goes *down*: rather than reacting to the reduction of jobs by creating new jobs, the consumer base reacts to the increase in cost by not being able to financially support the wages of *even* *more* *jobs*.
In 1790, 90% of Americans laborers (in a ~58% labor force) were farmers; we've replaced most of them with machines, and they now make up 2% of the labor force, and about 11% of consumer spending in total goes toward food to cover those farmers, the people building and maintaining farm equipment, logistics and sales moving that kind of thing, chemical companies making fertilizers and pesticides, and oil mining and refining to get the fuel for energy to drive all this. That means 18% of the labor involved in making food is on the farm, and 82% is in supporting infrastructure. You'll notice we don't have an 82% unemployment rate today; and automated fast food won't destroy our job market unless the method by which we transition is damaging--which this particular method *is*.
It doesn't fail as a fitness device. They acknowledge that spot-measurement of heart rate may be inaccurate; and also cite that such inaccuracies--even severe ones--don't affect its ability to use HRM data to calculate calories burned because it uses that data to measure statistical trends. In other words: FitBit isn't measuring (b) BPM and correlating that to (c) calories; it's measuring increases, decreases, and stable levels in heart rate, and determining if you're exerting yourself or not, and by how much (how rapidly and stably is your heart rate rising or falling?).
I've used FitBit and the FitBit Aria scale (which is, seriously, inaccurate with its body fat percentage measurement) along with calorie counting, and noticed a long-term trend of weight loss if I burned more calories than I took in. Weight fluctuates such that I might weigh 2 pounds more one day than the next, although I've held it pretty stable (half a pound fluctuation) by measuring at the same time under the same conditions (notably, in the morning, before eating, before showering, after pissing--minimal hydration level and post-fast); however, I lost 20 pounds in 3 months with moderate activity (walking for 20 minutes per day after lunch). I ate more when I burned more calories, and made sure I was coming under most of the time. As long as I kept that trend--even when I was burning 2700kcal/day and eating 2500, notably Popeye's Fried Chicken 1,100kcal lunch 3 days per week--my gut fat went down and my weight decreased.
Based on that, I would say they got the general trend correct, surprisingly so.