You're obviously responding to a non-technical person who has a vague understanding of a technical concept. GGPP would tend to use domain malapropisms that make sense in base-level English.
I gave you a domain bridge between "anti-social supernerd" and "normal people" in the hopes that you could communicate with regular human beings. I guess you're only interested in communicating with silicon and your cat.
The term "flag" in English has the generic definition of an artifact placed specifically as an indicator to modify the procedural strategy in process. Literal flags mark hazards and approved paths; conceptual flags include highlights and emphasis in text, markers on e-mails which need revisiting, compiler hints (e.g. likely() unlikely()), and CPU instruction code hints (branch prediction hints, prefetch/non-prefetch instructions, and so forth).
Modern branch predictors are highly complex; compilers rely on behaviors like loop reorganization to maximize cache hits, successful branch prediction, and out-of-order execution. Many of the hinting instructions have always been less-effective than using a conceptual model of the target CPU and reordering the code to maximize the effectiveness of its facilities. Modern JIT compiles (CIL, JVM) use information about the particular CPU architecture (a particular revision of an i7, an AMD CPU, etc.) to select for different sets of instructions to complete the same task, and to order them so as to maximize out-of-order execution, parallel execution, cache utilization, and successful branch prediction.
Actually, maternal time off work is covered by short- and long-term disability in the US. In the best setup, you get insurance that pays 66% of your salary, and you get your employer to make you pay for it from after-tax funds. If you use post-tax money, then claims come out of a conceptual taxed insurance pool: the people who pay with pre-tax or with employer money have to file claims as income and pay taxes, while the people who pay post-tax don't file claims as income.
That means if your employer pays the $5/paycheck for SLD, you get 66% of your salary minus the 35%-40% marginal total income tax (28% federal, 6.2% OASDI, 2%-6% state): you keep 39.6%-42.9% of your gross pay. If you pay it out of post-tax money, you might actually take home more from disability than your regular paycheck.
Six weeks of unpaid leave. Your employer doesn't pay shit. You just had a baby and the FMLA says you can't be fired for taking maternal leave. The whole time, your disability insurance pays you 66% of your gross pay, while your normal take-home pay is 65% of your gross pay; and, provided you paid taxes on the $5-per-paycheck premium, you don't pay taxes on the disability claim.
Take a year off work and your employer stops paying you, too.
There's no free money in the model. Your employer lets you transfer their risk, and they don't pay a penalty for your woman-habits. Whatever other argument you want to make about a pay gap, employers compensating for all the extra money paid to women who aren't actually working is bullshit.
Competence and productivity are immeasurable, especially at the point hiring decisions are made.
Look at my career: I have a CV that's impressive in some respects, but lagging in others. My employment history, achievements, and adaptability are clear; yet my CV doesn't carry the incredible weight of high-power, specialized technicians in computer security or systems management. In practice, there are trivialities I simply stall on because of gaps in my knowledge and a poor work ethic in specific situations; there are also insanely complex problems nobody else can solve as efficiently or effectively as I can, simply because I can effectively use analogical thinking and draw from an enormous source of broad and deep knowledge on a variety of topics to immediately comprehend complex systems made up of familiar or vaguely-familiar parts. I fall down when I hit a black box with unknown inputs and outputs.
That means not only does my CV not adequately describe my competence or productivity, but you can't adequately predict my competence and productivity in practice. I can perform poorly, average, or extremely well on any given problem; and most of the problems that come my way are new, which means I have to use old knowledge to shape out a new machine made of rearranged parts. I'm constantly grinding open black boxes, and also just flat-out failing on them. I figured out Puppet, Docker, and C# MVC; I can't get my head around OpenStack, Foreman, or all the front-end stuff in a Web application. I need someone to show me where the seams are so I can pry the black boxes open.
At hiring, I tend to get low-to-middling salaries, currently in the 50-percentile median as per Payscale. It's only by luck that it's worth it; and even then, I tend to replace all my job duties with heavy amounts of scripting and automation, systems that maintain themselves, and other labor-reducing solutions. I spend a lot of time getting paid to do nothing.
Is that competence? I have deep flaws in my competence in any practice. Is it productive? As long as you're not unlucky.
The rabbit hole gets deeper when you start factoring in things like ADHD (maybe that went away?), manic episodes, and other severe psychiatric problems. Good luck measuring the competency of someone who's crazy.
I've had RedHat break multiple subsystems during a point release. Once, they removed an entire configuration system (crmsh) and replaced it with a completely different configuration system. They outright removed the packages. I've seen them do major upgrades of sever software on a point release and put n the release notes that certain deprecated features are now non-features and your configuration won't work if you used them.
The weirdest one was when I upgraded httpd and Apache broke because you need to load a module to have ifenv, but the version that shipped in the prior point release had that module enabled by default. yum update resulted immediately in a broken web server, and the same config file that worked before didn't work anymore. I did that on three servers, though only the first one caught me off guard.
I'm not so nice: I switched to Ubuntu because, as much as I dislike their politics (Mir vs Wayland, Unity vs Gnome-3, etc.), they provide software *correctly*.
RHEL will suddenly break your shit mid-release. They won't ship an out-of-tree kernel module or a patch to a kernel driver to save your life, so good luck with RHEL 6 and Intel e1000e NICs (get the -lt kernel from ElRepo; but you can't remove RH Kernel or you break LSB, so have fun managing your bootloader as it keeps switching back to the broken one and then dropping network entirely on a kernel OOPS). They freeze the distro at a point release (6.7, 6.8, 6.9...) as they publish security patches, while ripping out some configuration subsystems and throwing in new ones (what has worked last week no longer works today, and you can freeze your release and not get further support!).
It's not even about being 9 years out-of-date on dot-zero release day--that's what Ubuntu over Debian is about. It's about having a zero-length support cycle on system-breaking changes. It's like if Microsoft upgraded Windows 2003 to Windows 2008 but kept calling it Windows 2003, and told you you could avoid all the major system changes by turning off all updates before that patch cycle--and suddenly the configuration files, registry entries, and supported features change or are outright deprecated and removed out from under you.
In Debian policy, you don't break things during release unless you have an extreme circumstance. If it worked on release day, it will work on EOL day. The software may be ass-old and out-of-date, but it's going to be the same software all throughout release, with the same configuration, the same features, and the same behavior. They might break something if that's the only way to remove a critical security vulnerability, but not if they can find another way around. Ubuntu turns this into a 6-month release cycle with 3 months to get your shit together after each release, or you can have a 2 year release cycle with 5 year support (LTS).
I never use RHEL-style distributions if I can avoid it.
They have a trademark for an instant-messaging application, not a JS library. Attempts to enforce the Kik trademark against NPM modules not dealing with instant messaging are vexatious, and any lawsuit raised is legal barratry.
Low-income areas in Washington, California (even around San Francisco), New York (western state especially, but some low-income areas in NYC), and really all over the United States can get as low as 60 cents per square foot; a $1/sqft model is reasonable as a reflection of the current low-income retail market everywhere.
In my Citizen's Dividend plan, I suggest 244sqft apartments at a $300/month budget in 2013. That's estimating $1.33/sqft, which is pretty high: a 610sqft place would be $811, and I've rented within 3/4 mile of two railway stations and even closer to a $200,000 art school for $725 for 750sqft.
People tell me it's undignified for a person to live in so little space; I tell them 600,000 Americans are living in alleyways, 4.8 million families (about 17 million Americans) are in homes only thanks to HUD assistance; and 48 million Americans go without enough food everyday. Dignity and an empty sack is worth the sack.
Who cares? We get access to the same stuff cheaper, and use the savings to access more stuff the rich used to have all to themselves. Your philosophical bullshit doesn't hold up much when food was 30% of the household income in 1950 and is 11% now, while cell phones went from $4,000 and $50/month plus 40 cents/minute in 1983 to $350 smart phone and $60 unlimited voice and data. By the way, that cell phone would be over $9,000 today, and you'd pay $550/month to talk for 2 hours per week.
People have no concept of intangible work. They see that shipping a digital game costs $0; what they don't consider is how much it costs to make a game. Oh, sure, the $9 million to make a 2D platformer in Unity is bullshit--I can do it in $100k or so ($58k for artwork! No kidding!)--but getting into high-end 3D game design with unique mechanics and large amounts of required testing *can* break the $1M barrier or more. That means you have to sell at least 20,000 copies ak $50 to break even, which might mean an extra quarter million or so in advertising, etc.
That doesn't include overhead costs of running a business, managing the projects (I used an idealized project run by a competent project manager, rather than a back-room project with a lot of rework, as a model), and making *other* games that have a negative return. If you make 10 games a year and 2 flop, all your games are 20% more expensive.
At some point all that employment is going to plateau for the people creating all the shit for people
It has and it does. That's what scarcity is, or at least one of the major effects.
Think about growing enough food to feed 12 people in 1 hectare of land (this is about 6 tonnes of rice per hectare, which is what India grew in 2000; in 1970, India could grow 2 tonnes per hectare). You need about 10 people full-time to work 4 hectares of land, and food costs roughly 2% of income as a result (add shipping and marketing and processing and such and you get the 13% of income).
So your population expands. 120 new people, 10 more hectares of land in use, 25 new people working that land, great. No change: more people working, more people buying, it's in the same proportion as current population, so population can grow like this FOREVER and will keep creating the same jobs in the same proportion.
Then you run out of arable land.
You have this dry, cracked land to work with. It's a bit harder. It's less fertile. You have to employ 10 people to work 4 hectares of land, plus 2 more people to make and supply fertilizer, plus 1 more person to supply irrigation. Even then, you get 50% as much yield from this land, so you can feed 6 people per hectare. You need about 13 people full-time to work 4 hectares of land, but that land feeds half as many people: to feed the original 48 people (previously from 4 hectares of land), you work 8 hectares with 26 people.
Food now costs more than twice as much to produce. For 120 new people, you put 20 more hectare of land in use, 65 people working full time. Notice this is 40 more people: where before you had 95 people to do other jobs, you only have 55 people to do other jobs. Even if we assume the consumer can buy all the same luxuries, nobody could make them to supply the consumer; and the idea that the consumer can buy them is ludicrous because the consumer must pay those 40 salaries to the food-makers. That makes consumers more poor.
This scarcity tends to raise the general cost of the scarce good, which can decrease the consumer market's broad buying power (everyone's food costs a little more, instead of the new kid's food costing a lot more), which can ultimately create a net-decrease in percentage employment. Other economic factors (notably, taxation of the middle class, and taxation on employment or sales) directly remove wage-labor income from consumers, thus factoring part of the cost of goods out of the buying power of consumers; those policy defects are much more powerful job-destroying factors, since scarcity of goods tends to move jobs and so doesn't have 100% destructive impact.
To use the GP's point, if a machine replaces 3 workers (that only requires 1 worker for machine maintenance), then that leaves 2 people unemployed. And your point is spot on here too; those 2 displaced workers will generally have to find work in another industry (services), as all the ex-farmers and ex-factory workers have had to do in the last 100 years or so.
Yes, and this becomes a problem if you lose workers too fast. If you pull an Industrial Revolution and unemploy 80% of your labor force in 5 years, you probably won't see new jobs.
The Information Age was interesting. We had a lot of information-heavy bottlenecks, like managing growing collections of legal contracts or accounting ledgers, and freeing that up actually uncorked the consumer machine. This contrasts strongly with the Industrial Revolution, which simply cut jobs without relieving scarcity.
Let's take American Express for example.
American Express originally checked *every* transaction through an Authorizer. The Authorizer would run a few quick checks and identify fraudulent transactions. If the transaction looked good, he'd allow it through, and everything was fabulous.
American Express could originally hire one (1) authorizer per 10,000 accounts. As American Express
You're thinking in terms of individual maximizing strategy; what about the whole market? Consumers have $1,000 to spend and need $200 to eat, but there are 6 people. To hire 1 person, you must pay him at least $200. To sell that product, you must charge at least $200. How many jobs can you create, and how many people are unemployed?
No amount of being in the right place at the right time will get that last person or two a job.
The answer is to make this happen more slowly by controlling employment costs.
A Dividend instead of a minimum wage eliminates the cost consequence of employment attached to a minimum-standard-of-living policy. Rather than passing a policy (minimum wage raise) forcing employers to pay more per unit wage-labor time, you pass a policy moving money to the working class (in the case of a Dividend or other UBI scheme, to a universal class e.g. all Adults) from a general funding source (an earmarked flat income tax is the most stable). Employers retaining employees face no disadvantage resulting from the policy change, whereas a minimum wage increase creates an added cost per wage-hour for employees previously below the new minimum.
Moving payroll taxes to general income taxes also reduces the cost per wage-hour. The argument is similar to the minimum wage argument.
A progressive general fund tax system allows the high-income-earner marginal tax brackets to remain static while the income gap increases (by way of high-income earners and businesses keeping a greater portion of any increasing wealth--the poor get richer, the rich get richer *faster*), allowing a reduction of taxes on the working class. This allows for a reduction of wages without a reduction of take-home wages, an increase in take-home wages without an increase in wages, or a middle-ground.
All of these efforts reduce the wage-labor cost basis of products, allowing prices to fall lower than otherwise possible while retaining profit margins. This increases the buying power of the consumer, as the proportion of the consumer's income which actually survives to expenditure (i.e. the part left after taxes) is increased. That creates more jobs.
Such efforts allow lower wage-labor costs, providing an advantage over newer low-labor techniques. This encourages businesses to wait for a strategic entry point: they may pay a worker $8/hr rather than replace him with a machine for $7.50/hr if they think the machine will be a 30-year, multi-million-dollar investment that pays off on the TCO of $5.50/hr *next* year. Each businesses has a different risk appetite (the amount of cost they *want* to put on the table for a possible benefit) and risk threshold (the point at which they no longer accept a risk), and so extending these strategic periods causes a spreading of labor replacement: instead of everyone buying machines one year, they start replacing workers over a span of ten years.
At the same time, the lower wage-labor cost means consumer buying power has a lower threshold before it can create new jobs; and the higher efficiency of wage-to-buying-power (how much of your paycheck you actually keep) increases the rate at which consumer buying power reaches that threshold. Paying employees lower wages and allowing them to take a bigger portion of that wage home leaves them more capable of affording new goods, which means you can more quickly replace lost jobs, keeping up with the above technological obsolescence.
Turn your study of technological history into plans and strategies. We need them. Our economists are so backwards they actually believe a sales tax is a good idea.
You're wrong; we'd be at -27,000% employment. What defect of thinking would lead you to believe we've only eliminated 3/4 of the jobs that ever existed since the 70s? We eliminated half of them, and then created new jobs; and we then eliminated half of those and created new ones; and eliminated half of those. Over a decade, we eliminate more jobs than the entire workforce population.
People don't find higher-wage employment by magic; we create new lower-wage jobs to replace lost jobs first. We can only buy new employment using the consumer dollars saved by eliminating existing jobs. The fact that those wages now can buy more stuff causes population to expand (until scarcity starts setting in again), and also allows us to pay wages representing an even lower share of the total income and still give people the same buying power.
You're right, though: The number of people needed to do a job doesn't increase. That's the point: technology *decreases* the number of people required to do a job, freeing their labor time up for other tasks. That's why we've moved people out of agriculture and manufacture and into construction, medicine, retail, and business services. Somebody has to sell those products from China; somebody has to handle the logistics, the distribution, the shipping; somebody has to drive the trucks; somebody has to run the involved IT systems.
Even after we've reduced the share of labor per product in *all* of these types of jobs, we create more jobs by buying more products. You buy 3 times as much shit, you need 3 times as much logistics. Maybe it takes 1/5 as much labor to provide those logistics, so you have 60% as many people doing that; the other 40% are running Spotify and Netflix.
We don't create higher-class jobs; we reduce costs and improve the standard-of-living of the lowest income earners. We may create more or fewer poor people; those poor people will be objectively wealthier than last generation's poor people, but they're still poor because literally everyone else has more than they do. Some of the replacement jobs are higher-income-class, some are lower-income-class, and we wind up with more things produced per wage-labor hour, more stuff per-capita, and more luxuries in the hands of everyone as their basic needs become cheaper.
If you replace people slowly enough, the unemployment gain comes more slowly than the replacement jobs created by the increased consumer buying power. Unemployment remains stable, goods become cheaper, people find new jobs, and everyone becomes wealthier. You can improve this by keeping labor cheap, notably by using alternate strategies to minimum wage (e.g. a non-wage Citizen's Dividend), replacing payroll taxes with income taxes, eliminating sales and VAT taxes, and reducing working-class income taxes (by using a progressive tax system and leaving the high-income tax brackets static while lowering the low-income marginal taxes as the income gap widens).
If you do it quickly enough, you create massive unemployment and get the Industrial Revolution.
We have the Luddites, who believe these changes eliminate jobs *forever*; and the Technocrats, who believe new technology creates new jobs immediately. A lot of the Basic Income dialogue isn't about slowing and steadying the movement onto new technology and keeping people employable so we can create new jobs faster; it's about saving us when we make employment an obsolete concept. A lot of people also think we should raise minimum wage to *encourage* businesses to move onto machines, since that will somehow bring us prosperity by making things cheaper; they ignore that they're trying to make people to move onto machines that are more expensive than modern methods, thus making things more expensive.
Lots of people with no concept of economics. Even economists have no concept of economics. I have made a *lot* of economic discourse to explain why supply-and-demand works or how technology affects employment to people who have degrees in economics because modern theory can spot things like this as effects, but can't identify the mechanism. We have people with Ph.D.s in economics arguing over whether automation is going to cause one thing or the other, and I'm over here like, "You're both right, and you're both wrong; your models track the effect, but not the cause."
Imagine what it's like dealing with armchair economists who read something about supply-and-demand on Wikipedia, but have never taken the time to study economic history and work out any behavioral models. Every other word is "competition", and they keep babbling about an economy getting wealthier by the same dollar being spent over and over again rapidly.
A passphrase is inadequate. It's like you said, "I have a treasure chest filled with soled gold in my front yard; I should put a $2 padlock for a girl's diary on it to protect it." Yeah, I'll just pull harder on the latch; that lock will pop with just the force of my bare hands, no tools needed.
"Better than nothing" is not a strong vote of confidence. A seat belt is apparently better than nothing in a land speed record attempt, but you're still going to disintegrate into a cloud of red dust if you hit the tiniest bump.
We have unemployment because unemployment is unavoidable. You having a job means somewhere, someone else doesn't have that job; the tube holds 30 balls and, eventually, putting one in one side pops another out the other side.
The length of the tube is determined by the consumer market's ability to buy. Whether or not you get in the tube is determined by luck first, and then your own actions; you can wind up jobless despite your heroic effort.
It is relevant to those who are accused of anything. We can't use your giant media collection as criminal evidence because it's found under search for a different thing; but the discovery can go to the RIAA, who then claims you don't have proof that you ever owned any of these CDs, and thus must have acquired these files illegally. Billion dollar lawsuit.
It's not so much that we're facing more serious issues as that this isn't a real issue. They're trying to make an application function optimally for the user, including acting exactly as it does with an uncapped pipe *and* getting the user's cost down. It's like you're buying beer and carrying it in a bucket with a hole in the bottom, and they've tried to plug the hole--which makes it a poor flower pot (no drainage), but you're not using that particular bucket as a flower pot. There might or might not be some corner cases.
I'm not big on the argument that something is okay because there are worse things. I like well-engineered solutions, not excuses about how something can be bad because something else is twice as bad. The "Binge On" policy is a well-engineered solution; it might have some flaws, and better engineering will fix those (again: no excuse for being bad just because something else is worse--unless that something else is the best available alternative).
The fact is HD video targeting a 5-inch OLED cell phone screen only needs 1.5Mbit/s at 1080p h.264; h.265 can apparently get as small as half the bitrate for the same quality.
If you try to view a 3 minute video, watch the first 15 seconds, and buffer 2 minutes across your 12Mbit/s connection, you've spent 15 seconds holding the pipe at a high data rate only to discard all that data. With 100 users, that's got to be a 1.2Gbit pipe, and so you have to provide that kind of service with the cost divided among 100 users.
If the data is throttled to 1.5Mbit/s, suddenly you can provide that kind of streaming video service to 800 users at once. They can each pay 1/8 as much of the cost, and they all receive the same service they were trying to get anyway. If they switch to something other than video, you have to give them the full 12Mbit/s; but if they're watching Netflix, you can cut back their pipe for that particular stream.
T-Mobile has observed that this makes it reasonable to just *not* meter data if they throttle the pipe while people watch Netflix. The more of their video traffic they swing under this strategy, the more costs they save; and the end result *should* be transparent to the user. There have been a few bugs--the download bitrate is throttled if you try to download a video file because the DPI is too dumb to recognize you're not streaming--but it mostly works as advertised.
So long as they make reasonable effort to supply the same service as an unthrottled connection--you can still watch as many Netflix or Youtube streams (on a phone, that's ONE) as with generic service--they haven't slowed down access or created a premium service; they've ensured a service works at minimal cost, maximizing efficiency and reducing the cost to the consumer.
Broken is still broken, but it's not an evil conspiracy. Someone will still say they're trying to cut cost and screw the consumer while ignoring that they outright stop charging the consumer for the data.
That's true, and it's basic entrepreneurship; I was more commenting on the more basic premise that you can't sell any sort of new product in a market where consumers can't pay the wages of everybody involved in getting that product made, shipped, and sold. If the consumer base can't pay for the new jobs, then you can't create new jobs, and thus *somebody* has to lose.
If all small businesses succeeded, established businesses would shrink and die constantly.
Most people don't know anything about economics. Even economists don't really know that much about economics--something I realized when I started looking at actual modern economic theory and found only naive observations and general rules with no understanding of mechanism.
Take a look at the job creation narrative. We've seen all kinds of talk about things like trickle-down economics (which actual economists know is bullshit--let's be fair: it's pundits and politicians who believe that crap), leading to this ideal where you give businesses income tax cuts and they create jobs.... How? Even a cursory examination of market economics in a macroeconomic context says that's nonsense--enough nonsense that you can check for the fallacy of common sense and verify that, yes, the sensible conclusion *is* sensible and the one about trickle-down is ludicrous.
Businesses aren't charities and don't create jobs out of excess profits just so some bloke can get by. Common sense; but this is not sufficient because common sense is often wrong and stupid.
Businesses don't pay wages.
Every product has a cost, which ultimately comes down to wage-labor: in aggregate, the shelf price of a product is its wage-labor cost plus some profit margin. You pay your employees's wages, plus payroll tax, plus benefits; if you charge less than the fraction of these wage-labor costs involved in making a product, you run at a net loss, and go insolvent. That means the minimum sustainable price--the price the consumer must pay--is wages. When GM negotiates for 100 million tonnes of steel per year for 10 years to make cars, the steelmaker goes back to multiple coal and ore suppliers and does the same, and everyone starts competing by lowering big profit margins closer to *and* *never* *less* *than* their costs--which, down the whole supply chain, comes down to squeezing out the aggregate profit margins and bringing the price of steel closer to the cost to a vertical monopoly owning the mines, fuel supplies, and all production mills. It can't go any lower.
You can't supply a product consumers can't afford, which means the consumer base making up the product's market has to be able to pay all of the wages involved in making, moving, and selling the product. The manufacturing wages have to be paid; the transportation wages have to be paid; the logistics wages have to be paid; the retail wages have to be paid. All kinds of taxes are involved. After that, you slap on some mark-up that a sufficiently-sized consumer market will tolerate (and can afford) and you make a profit. If the consumer can't afford it, none of those jobs come into existence.
So we have senators talking out one side of their mouths about "Job Creation", and out the other side they're talking about levying sales taxes (a particularly destructive way to artificially raise the cost-basis of a product to the consumer). We have Social Security OASDI applied to the broad consumer base as a 12.4% income tax--6.2% levied on your paycheck and 6.2% levied on PAYROLL (holy shit that's daft)--rather than as a
dedicated flat or progressive income tax. They find every way to destroy jobs while talking about the need to create new jobs.
These same people start talking about how they'll get us all free college so we have the ability to go find or *make* jobs for ourselves. They talk about small businesses and the American Dream, and draw up a narrative about how jobs come from hard work and dedication, and not from capturing the limited supply of consumer money.
It's not that they're all evil conspirators trying to distort and manipulate our economy; it's that they don't know, they don't understand. They think they have it right; they're just wrong.
When someone tells you small businesses are failing like crazy, think about where jobs come from.
Comcast has been doing this nation-wide for a few years now. Their "Blast!" service (as in, "Blast! It's out again!", although that seems to happen a lot at 2 in the morning...) has stepped up from 50Mbit/s to 80Mbit/s to 110Mbit/s to now 150Mbit/s; the day Blast! got re-labeled to 150Mbit/s, it was clocking in real speeds of 179Mbit/s. The price has been $79/month plus like $11 in taxes and fees (because it wouldn't be Comcast without horse shit) for nearly a decade.
People have been talking about overbuilt Internet infrastructure and dark fiber for years. Comcast and Verizon are suddenly rolling out all this high-speed service. They realized they're not monetizing any of this shit anyway, and they're having a PR arms race as the equipment to actually use it becomes cheaper--that is: the arms race has become sustainable for the long term, so they're not risking a costly upgrade or an embarrassing PR fumble by doing this now.
Slashdot Mirror
You're obviously responding to a non-technical person who has a vague understanding of a technical concept. GGPP would tend to use domain malapropisms that make sense in base-level English.
I gave you a domain bridge between "anti-social supernerd" and "normal people" in the hopes that you could communicate with regular human beings. I guess you're only interested in communicating with silicon and your cat.
The term "flag" in English has the generic definition of an artifact placed specifically as an indicator to modify the procedural strategy in process. Literal flags mark hazards and approved paths; conceptual flags include highlights and emphasis in text, markers on e-mails which need revisiting, compiler hints (e.g. likely() unlikely()), and CPU instruction code hints (branch prediction hints, prefetch/non-prefetch instructions, and so forth).
Modern branch predictors are highly complex; compilers rely on behaviors like loop reorganization to maximize cache hits, successful branch prediction, and out-of-order execution. Many of the hinting instructions have always been less-effective than using a conceptual model of the target CPU and reordering the code to maximize the effectiveness of its facilities. Modern JIT compiles (CIL, JVM) use information about the particular CPU architecture (a particular revision of an i7, an AMD CPU, etc.) to select for different sets of instructions to complete the same task, and to order them so as to maximize out-of-order execution, parallel execution, cache utilization, and successful branch prediction.
Actually, maternal time off work is covered by short- and long-term disability in the US. In the best setup, you get insurance that pays 66% of your salary, and you get your employer to make you pay for it from after-tax funds. If you use post-tax money, then claims come out of a conceptual taxed insurance pool: the people who pay with pre-tax or with employer money have to file claims as income and pay taxes, while the people who pay post-tax don't file claims as income.
That means if your employer pays the $5/paycheck for SLD, you get 66% of your salary minus the 35%-40% marginal total income tax (28% federal, 6.2% OASDI, 2%-6% state): you keep 39.6%-42.9% of your gross pay. If you pay it out of post-tax money, you might actually take home more from disability than your regular paycheck.
Six weeks of unpaid leave. Your employer doesn't pay shit. You just had a baby and the FMLA says you can't be fired for taking maternal leave. The whole time, your disability insurance pays you 66% of your gross pay, while your normal take-home pay is 65% of your gross pay; and, provided you paid taxes on the $5-per-paycheck premium, you don't pay taxes on the disability claim.
Take a year off work and your employer stops paying you, too.
There's no free money in the model. Your employer lets you transfer their risk, and they don't pay a penalty for your woman-habits. Whatever other argument you want to make about a pay gap, employers compensating for all the extra money paid to women who aren't actually working is bullshit.
Competence and productivity are immeasurable, especially at the point hiring decisions are made.
Look at my career: I have a CV that's impressive in some respects, but lagging in others. My employment history, achievements, and adaptability are clear; yet my CV doesn't carry the incredible weight of high-power, specialized technicians in computer security or systems management. In practice, there are trivialities I simply stall on because of gaps in my knowledge and a poor work ethic in specific situations; there are also insanely complex problems nobody else can solve as efficiently or effectively as I can, simply because I can effectively use analogical thinking and draw from an enormous source of broad and deep knowledge on a variety of topics to immediately comprehend complex systems made up of familiar or vaguely-familiar parts. I fall down when I hit a black box with unknown inputs and outputs.
That means not only does my CV not adequately describe my competence or productivity, but you can't adequately predict my competence and productivity in practice. I can perform poorly, average, or extremely well on any given problem; and most of the problems that come my way are new, which means I have to use old knowledge to shape out a new machine made of rearranged parts. I'm constantly grinding open black boxes, and also just flat-out failing on them. I figured out Puppet, Docker, and C# MVC; I can't get my head around OpenStack, Foreman, or all the front-end stuff in a Web application. I need someone to show me where the seams are so I can pry the black boxes open.
At hiring, I tend to get low-to-middling salaries, currently in the 50-percentile median as per Payscale. It's only by luck that it's worth it; and even then, I tend to replace all my job duties with heavy amounts of scripting and automation, systems that maintain themselves, and other labor-reducing solutions. I spend a lot of time getting paid to do nothing.
Is that competence? I have deep flaws in my competence in any practice. Is it productive? As long as you're not unlucky.
The rabbit hole gets deeper when you start factoring in things like ADHD (maybe that went away?), manic episodes, and other severe psychiatric problems. Good luck measuring the competency of someone who's crazy.
I've had RedHat break multiple subsystems during a point release. Once, they removed an entire configuration system (crmsh) and replaced it with a completely different configuration system. They outright removed the packages. I've seen them do major upgrades of sever software on a point release and put n the release notes that certain deprecated features are now non-features and your configuration won't work if you used them.
The weirdest one was when I upgraded httpd and Apache broke because you need to load a module to have ifenv, but the version that shipped in the prior point release had that module enabled by default. yum update resulted immediately in a broken web server, and the same config file that worked before didn't work anymore. I did that on three servers, though only the first one caught me off guard.
I'm not so nice: I switched to Ubuntu because, as much as I dislike their politics (Mir vs Wayland, Unity vs Gnome-3, etc.), they provide software *correctly*.
RHEL will suddenly break your shit mid-release. They won't ship an out-of-tree kernel module or a patch to a kernel driver to save your life, so good luck with RHEL 6 and Intel e1000e NICs (get the -lt kernel from ElRepo; but you can't remove RH Kernel or you break LSB, so have fun managing your bootloader as it keeps switching back to the broken one and then dropping network entirely on a kernel OOPS). They freeze the distro at a point release (6.7, 6.8, 6.9...) as they publish security patches, while ripping out some configuration subsystems and throwing in new ones (what has worked last week no longer works today, and you can freeze your release and not get further support!).
It's not even about being 9 years out-of-date on dot-zero release day--that's what Ubuntu over Debian is about. It's about having a zero-length support cycle on system-breaking changes. It's like if Microsoft upgraded Windows 2003 to Windows 2008 but kept calling it Windows 2003, and told you you could avoid all the major system changes by turning off all updates before that patch cycle--and suddenly the configuration files, registry entries, and supported features change or are outright deprecated and removed out from under you.
In Debian policy, you don't break things during release unless you have an extreme circumstance. If it worked on release day, it will work on EOL day. The software may be ass-old and out-of-date, but it's going to be the same software all throughout release, with the same configuration, the same features, and the same behavior. They might break something if that's the only way to remove a critical security vulnerability, but not if they can find another way around. Ubuntu turns this into a 6-month release cycle with 3 months to get your shit together after each release, or you can have a 2 year release cycle with 5 year support (LTS).
I never use RHEL-style distributions if I can avoid it.
They have a trademark for an instant-messaging application, not a JS library. Attempts to enforce the Kik trademark against NPM modules not dealing with instant messaging are vexatious, and any lawsuit raised is legal barratry.
Low-income areas in Washington, California (even around San Francisco), New York (western state especially, but some low-income areas in NYC), and really all over the United States can get as low as 60 cents per square foot; a $1/sqft model is reasonable as a reflection of the current low-income retail market everywhere.
In my Citizen's Dividend plan, I suggest 244sqft apartments at a $300/month budget in 2013. That's estimating $1.33/sqft, which is pretty high: a 610sqft place would be $811, and I've rented within 3/4 mile of two railway stations and even closer to a $200,000 art school for $725 for 750sqft.
People tell me it's undignified for a person to live in so little space; I tell them 600,000 Americans are living in alleyways, 4.8 million families (about 17 million Americans) are in homes only thanks to HUD assistance; and 48 million Americans go without enough food everyday. Dignity and an empty sack is worth the sack.
Who cares? We get access to the same stuff cheaper, and use the savings to access more stuff the rich used to have all to themselves. Your philosophical bullshit doesn't hold up much when food was 30% of the household income in 1950 and is 11% now, while cell phones went from $4,000 and $50/month plus 40 cents/minute in 1983 to $350 smart phone and $60 unlimited voice and data. By the way, that cell phone would be over $9,000 today, and you'd pay $550/month to talk for 2 hours per week.
People have no concept of intangible work. They see that shipping a digital game costs $0; what they don't consider is how much it costs to make a game. Oh, sure, the $9 million to make a 2D platformer in Unity is bullshit--I can do it in $100k or so ($58k for artwork! No kidding!)--but getting into high-end 3D game design with unique mechanics and large amounts of required testing *can* break the $1M barrier or more. That means you have to sell at least 20,000 copies ak $50 to break even, which might mean an extra quarter million or so in advertising, etc.
That doesn't include overhead costs of running a business, managing the projects (I used an idealized project run by a competent project manager, rather than a back-room project with a lot of rework, as a model), and making *other* games that have a negative return. If you make 10 games a year and 2 flop, all your games are 20% more expensive.
We still get $15 games on Steam.
At some point all that employment is going to plateau for the people creating all the shit for people
It has and it does. That's what scarcity is, or at least one of the major effects.
Think about growing enough food to feed 12 people in 1 hectare of land (this is about 6 tonnes of rice per hectare, which is what India grew in 2000; in 1970, India could grow 2 tonnes per hectare). You need about 10 people full-time to work 4 hectares of land, and food costs roughly 2% of income as a result (add shipping and marketing and processing and such and you get the 13% of income).
So your population expands. 120 new people, 10 more hectares of land in use, 25 new people working that land, great. No change: more people working, more people buying, it's in the same proportion as current population, so population can grow like this FOREVER and will keep creating the same jobs in the same proportion.
Then you run out of arable land.
You have this dry, cracked land to work with. It's a bit harder. It's less fertile. You have to employ 10 people to work 4 hectares of land, plus 2 more people to make and supply fertilizer, plus 1 more person to supply irrigation. Even then, you get 50% as much yield from this land, so you can feed 6 people per hectare. You need about 13 people full-time to work 4 hectares of land, but that land feeds half as many people: to feed the original 48 people (previously from 4 hectares of land), you work 8 hectares with 26 people.
Food now costs more than twice as much to produce. For 120 new people, you put 20 more hectare of land in use, 65 people working full time. Notice this is 40 more people: where before you had 95 people to do other jobs, you only have 55 people to do other jobs. Even if we assume the consumer can buy all the same luxuries, nobody could make them to supply the consumer; and the idea that the consumer can buy them is ludicrous because the consumer must pay those 40 salaries to the food-makers. That makes consumers more poor.
This scarcity tends to raise the general cost of the scarce good, which can decrease the consumer market's broad buying power (everyone's food costs a little more, instead of the new kid's food costing a lot more), which can ultimately create a net-decrease in percentage employment. Other economic factors (notably, taxation of the middle class, and taxation on employment or sales) directly remove wage-labor income from consumers, thus factoring part of the cost of goods out of the buying power of consumers; those policy defects are much more powerful job-destroying factors, since scarcity of goods tends to move jobs and so doesn't have 100% destructive impact.
To use the GP's point, if a machine replaces 3 workers (that only requires 1 worker for machine maintenance), then that leaves 2 people unemployed. And your point is spot on here too; those 2 displaced workers will generally have to find work in another industry (services), as all the ex-farmers and ex-factory workers have had to do in the last 100 years or so.
Yes, and this becomes a problem if you lose workers too fast. If you pull an Industrial Revolution and unemploy 80% of your labor force in 5 years, you probably won't see new jobs.
The Information Age was interesting. We had a lot of information-heavy bottlenecks, like managing growing collections of legal contracts or accounting ledgers, and freeing that up actually uncorked the consumer machine. This contrasts strongly with the Industrial Revolution, which simply cut jobs without relieving scarcity.
Let's take American Express for example.
American Express originally checked *every* transaction through an Authorizer. The Authorizer would run a few quick checks and identify fraudulent transactions. If the transaction looked good, he'd allow it through, and everything was fabulous.
American Express could originally hire one (1) authorizer per 10,000 accounts. As American Express
You're thinking in terms of individual maximizing strategy; what about the whole market? Consumers have $1,000 to spend and need $200 to eat, but there are 6 people. To hire 1 person, you must pay him at least $200. To sell that product, you must charge at least $200. How many jobs can you create, and how many people are unemployed?
No amount of being in the right place at the right time will get that last person or two a job.
The answer is to make this happen more slowly by controlling employment costs.
A Dividend instead of a minimum wage eliminates the cost consequence of employment attached to a minimum-standard-of-living policy. Rather than passing a policy (minimum wage raise) forcing employers to pay more per unit wage-labor time, you pass a policy moving money to the working class (in the case of a Dividend or other UBI scheme, to a universal class e.g. all Adults) from a general funding source (an earmarked flat income tax is the most stable). Employers retaining employees face no disadvantage resulting from the policy change, whereas a minimum wage increase creates an added cost per wage-hour for employees previously below the new minimum.
Moving payroll taxes to general income taxes also reduces the cost per wage-hour. The argument is similar to the minimum wage argument.
A progressive general fund tax system allows the high-income-earner marginal tax brackets to remain static while the income gap increases (by way of high-income earners and businesses keeping a greater portion of any increasing wealth--the poor get richer, the rich get richer *faster*), allowing a reduction of taxes on the working class. This allows for a reduction of wages without a reduction of take-home wages, an increase in take-home wages without an increase in wages, or a middle-ground.
All of these efforts reduce the wage-labor cost basis of products, allowing prices to fall lower than otherwise possible while retaining profit margins. This increases the buying power of the consumer, as the proportion of the consumer's income which actually survives to expenditure (i.e. the part left after taxes) is increased. That creates more jobs.
Such efforts allow lower wage-labor costs, providing an advantage over newer low-labor techniques. This encourages businesses to wait for a strategic entry point: they may pay a worker $8/hr rather than replace him with a machine for $7.50/hr if they think the machine will be a 30-year, multi-million-dollar investment that pays off on the TCO of $5.50/hr *next* year. Each businesses has a different risk appetite (the amount of cost they *want* to put on the table for a possible benefit) and risk threshold (the point at which they no longer accept a risk), and so extending these strategic periods causes a spreading of labor replacement: instead of everyone buying machines one year, they start replacing workers over a span of ten years.
At the same time, the lower wage-labor cost means consumer buying power has a lower threshold before it can create new jobs; and the higher efficiency of wage-to-buying-power (how much of your paycheck you actually keep) increases the rate at which consumer buying power reaches that threshold. Paying employees lower wages and allowing them to take a bigger portion of that wage home leaves them more capable of affording new goods, which means you can more quickly replace lost jobs, keeping up with the above technological obsolescence.
Turn your study of technological history into plans and strategies. We need them. Our economists are so backwards they actually believe a sales tax is a good idea.
You're wrong; we'd be at -27,000% employment. What defect of thinking would lead you to believe we've only eliminated 3/4 of the jobs that ever existed since the 70s? We eliminated half of them, and then created new jobs; and we then eliminated half of those and created new ones; and eliminated half of those. Over a decade, we eliminate more jobs than the entire workforce population.
People don't find higher-wage employment by magic; we create new lower-wage jobs to replace lost jobs first. We can only buy new employment using the consumer dollars saved by eliminating existing jobs. The fact that those wages now can buy more stuff causes population to expand (until scarcity starts setting in again), and also allows us to pay wages representing an even lower share of the total income and still give people the same buying power.
We didn't have cell phones in the 1700s.
There's population versus technology, there's people spending less on food and clothes while buying bigger houses, the increase in non-necessity expenditures as a percentage of income, and all kinds of other data showing we actually do increase the number of people in the economy.
You're right, though: The number of people needed to do a job doesn't increase. That's the point: technology *decreases* the number of people required to do a job, freeing their labor time up for other tasks. That's why we've moved people out of agriculture and manufacture and into construction, medicine, retail, and business services. Somebody has to sell those products from China; somebody has to handle the logistics, the distribution, the shipping; somebody has to drive the trucks; somebody has to run the involved IT systems.
Even after we've reduced the share of labor per product in *all* of these types of jobs, we create more jobs by buying more products. You buy 3 times as much shit, you need 3 times as much logistics. Maybe it takes 1/5 as much labor to provide those logistics, so you have 60% as many people doing that; the other 40% are running Spotify and Netflix.
We don't create higher-class jobs; we reduce costs and improve the standard-of-living of the lowest income earners. We may create more or fewer poor people; those poor people will be objectively wealthier than last generation's poor people, but they're still poor because literally everyone else has more than they do. Some of the replacement jobs are higher-income-class, some are lower-income-class, and we wind up with more things produced per wage-labor hour, more stuff per-capita, and more luxuries in the hands of everyone as their basic needs become cheaper.
If you replace people slowly enough, the unemployment gain comes more slowly than the replacement jobs created by the increased consumer buying power. Unemployment remains stable, goods become cheaper, people find new jobs, and everyone becomes wealthier. You can improve this by keeping labor cheap, notably by using alternate strategies to minimum wage (e.g. a non-wage Citizen's Dividend), replacing payroll taxes with income taxes, eliminating sales and VAT taxes, and reducing working-class income taxes (by using a progressive tax system and leaving the high-income tax brackets static while lowering the low-income marginal taxes as the income gap widens).
If you do it quickly enough, you create massive unemployment and get the Industrial Revolution.
We have the Luddites, who believe these changes eliminate jobs *forever*; and the Technocrats, who believe new technology creates new jobs immediately. A lot of the Basic Income dialogue isn't about slowing and steadying the movement onto new technology and keeping people employable so we can create new jobs faster; it's about saving us when we make employment an obsolete concept. A lot of people also think we should raise minimum wage to *encourage* businesses to move onto machines, since that will somehow bring us prosperity by making things cheaper; they ignore that they're trying to make people to move onto machines that are more expensive than modern methods, thus making things more expensive.
Lots of people with no concept of economics. Even economists have no concept of economics. I have made a *lot* of economic discourse to explain why supply-and-demand works or how technology affects employment to people who have degrees in economics because modern theory can spot things like this as effects, but can't identify the mechanism. We have people with Ph.D.s in economics arguing over whether automation is going to cause one thing or the other, and I'm over here like, "You're both right, and you're both wrong; your models track the effect, but not the cause."
Imagine what it's like dealing with armchair economists who read something about supply-and-demand on Wikipedia, but have never taken the time to study economic history and work out any behavioral models. Every other word is "competition", and they keep babbling about an economy getting wealthier by the same dollar being spent over and over again rapidly.
A passphrase is inadequate. It's like you said, "I have a treasure chest filled with soled gold in my front yard; I should put a $2 padlock for a girl's diary on it to protect it." Yeah, I'll just pull harder on the latch; that lock will pop with just the force of my bare hands, no tools needed.
"Better than nothing" is not a strong vote of confidence. A seat belt is apparently better than nothing in a land speed record attempt, but you're still going to disintegrate into a cloud of red dust if you hit the tiniest bump.
We have unemployment because unemployment is unavoidable. You having a job means somewhere, someone else doesn't have that job; the tube holds 30 balls and, eventually, putting one in one side pops another out the other side.
The length of the tube is determined by the consumer market's ability to buy. Whether or not you get in the tube is determined by luck first, and then your own actions; you can wind up jobless despite your heroic effort.
It is relevant to those who are accused of anything. We can't use your giant media collection as criminal evidence because it's found under search for a different thing; but the discovery can go to the RIAA, who then claims you don't have proof that you ever owned any of these CDs, and thus must have acquired these files illegally. Billion dollar lawsuit.
Passphrase-encrypted means unlimited-rate cracking.
It's not so much that we're facing more serious issues as that this isn't a real issue. They're trying to make an application function optimally for the user, including acting exactly as it does with an uncapped pipe *and* getting the user's cost down. It's like you're buying beer and carrying it in a bucket with a hole in the bottom, and they've tried to plug the hole--which makes it a poor flower pot (no drainage), but you're not using that particular bucket as a flower pot. There might or might not be some corner cases.
I'm not big on the argument that something is okay because there are worse things. I like well-engineered solutions, not excuses about how something can be bad because something else is twice as bad. The "Binge On" policy is a well-engineered solution; it might have some flaws, and better engineering will fix those (again: no excuse for being bad just because something else is worse--unless that something else is the best available alternative).
The fact is HD video targeting a 5-inch OLED cell phone screen only needs 1.5Mbit/s at 1080p h.264; h.265 can apparently get as small as half the bitrate for the same quality.
If you try to view a 3 minute video, watch the first 15 seconds, and buffer 2 minutes across your 12Mbit/s connection, you've spent 15 seconds holding the pipe at a high data rate only to discard all that data. With 100 users, that's got to be a 1.2Gbit pipe, and so you have to provide that kind of service with the cost divided among 100 users.
If the data is throttled to 1.5Mbit/s, suddenly you can provide that kind of streaming video service to 800 users at once. They can each pay 1/8 as much of the cost, and they all receive the same service they were trying to get anyway. If they switch to something other than video, you have to give them the full 12Mbit/s; but if they're watching Netflix, you can cut back their pipe for that particular stream.
T-Mobile has observed that this makes it reasonable to just *not* meter data if they throttle the pipe while people watch Netflix. The more of their video traffic they swing under this strategy, the more costs they save; and the end result *should* be transparent to the user. There have been a few bugs--the download bitrate is throttled if you try to download a video file because the DPI is too dumb to recognize you're not streaming--but it mostly works as advertised.
So long as they make reasonable effort to supply the same service as an unthrottled connection--you can still watch as many Netflix or Youtube streams (on a phone, that's ONE) as with generic service--they haven't slowed down access or created a premium service; they've ensured a service works at minimal cost, maximizing efficiency and reducing the cost to the consumer.
Broken is still broken, but it's not an evil conspiracy. Someone will still say they're trying to cut cost and screw the consumer while ignoring that they outright stop charging the consumer for the data.
That's true, and it's basic entrepreneurship; I was more commenting on the more basic premise that you can't sell any sort of new product in a market where consumers can't pay the wages of everybody involved in getting that product made, shipped, and sold. If the consumer base can't pay for the new jobs, then you can't create new jobs, and thus *somebody* has to lose.
If all small businesses succeeded, established businesses would shrink and die constantly.
Most people don't know anything about economics. Even economists don't really know that much about economics--something I realized when I started looking at actual modern economic theory and found only naive observations and general rules with no understanding of mechanism.
Take a look at the job creation narrative. We've seen all kinds of talk about things like trickle-down economics (which actual economists know is bullshit--let's be fair: it's pundits and politicians who believe that crap), leading to this ideal where you give businesses income tax cuts and they create jobs. ... How? Even a cursory examination of market economics in a macroeconomic context says that's nonsense--enough nonsense that you can check for the fallacy of common sense and verify that, yes, the sensible conclusion *is* sensible and the one about trickle-down is ludicrous.
Businesses aren't charities and don't create jobs out of excess profits just so some bloke can get by. Common sense; but this is not sufficient because common sense is often wrong and stupid.
Businesses don't pay wages.
Every product has a cost, which ultimately comes down to wage-labor: in aggregate, the shelf price of a product is its wage-labor cost plus some profit margin. You pay your employees's wages, plus payroll tax, plus benefits; if you charge less than the fraction of these wage-labor costs involved in making a product, you run at a net loss, and go insolvent. That means the minimum sustainable price--the price the consumer must pay--is wages. When GM negotiates for 100 million tonnes of steel per year for 10 years to make cars, the steelmaker goes back to multiple coal and ore suppliers and does the same, and everyone starts competing by lowering big profit margins closer to *and* *never* *less* *than* their costs--which, down the whole supply chain, comes down to squeezing out the aggregate profit margins and bringing the price of steel closer to the cost to a vertical monopoly owning the mines, fuel supplies, and all production mills. It can't go any lower.
You can't supply a product consumers can't afford, which means the consumer base making up the product's market has to be able to pay all of the wages involved in making, moving, and selling the product. The manufacturing wages have to be paid; the transportation wages have to be paid; the logistics wages have to be paid; the retail wages have to be paid. All kinds of taxes are involved. After that, you slap on some mark-up that a sufficiently-sized consumer market will tolerate (and can afford) and you make a profit. If the consumer can't afford it, none of those jobs come into existence.
So we have senators talking out one side of their mouths about "Job Creation", and out the other side they're talking about levying sales taxes (a particularly destructive way to artificially raise the cost-basis of a product to the consumer). We have Social Security OASDI applied to the broad consumer base as a 12.4% income tax--6.2% levied on your paycheck and 6.2% levied on PAYROLL (holy shit that's daft)--rather than as a dedicated flat or progressive income tax. They find every way to destroy jobs while talking about the need to create new jobs.
These same people start talking about how they'll get us all free college so we have the ability to go find or *make* jobs for ourselves. They talk about small businesses and the American Dream, and draw up a narrative about how jobs come from hard work and dedication, and not from capturing the limited supply of consumer money.
It's not that they're all evil conspirators trying to distort and manipulate our economy; it's that they don't know, they don't understand. They think they have it right; they're just wrong.
When someone tells you small businesses are failing like crazy, think about where jobs come from.
Comcast has been doing this nation-wide for a few years now. Their "Blast!" service (as in, "Blast! It's out again!", although that seems to happen a lot at 2 in the morning...) has stepped up from 50Mbit/s to 80Mbit/s to 110Mbit/s to now 150Mbit/s; the day Blast! got re-labeled to 150Mbit/s, it was clocking in real speeds of 179Mbit/s. The price has been $79/month plus like $11 in taxes and fees (because it wouldn't be Comcast without horse shit) for nearly a decade.
People have been talking about overbuilt Internet infrastructure and dark fiber for years. Comcast and Verizon are suddenly rolling out all this high-speed service. They realized they're not monetizing any of this shit anyway, and they're having a PR arms race as the equipment to actually use it becomes cheaper--that is: the arms race has become sustainable for the long term, so they're not risking a costly upgrade or an embarrassing PR fumble by doing this now.