Many jobs can be automated, but not be cost effective. I imagine that as the cost of the fast food worker rises, for instance, the research on replacing that worker with a robot will also increase. It will be seen if robots are tolerated in what right now is a face to face encounter.

The telemarketer has already been replaced by robots, but robots are not tolerated so these jobs are still secure. It is the same reason that these jobs are still present in the US instead of completely exported to other countries. Consumer demand.

I still think that lawyers are doctors are going to see the greatest impact in wages and jobs. The salaries for first year lawyers, for example, have been fixed or falling for a decade according to published reports. As more data is collected on patients, and that data is correlated to outcomes, the heuristics and stochastic will reach a level where only the best diagnosticians will remain employable.

Here is how i interpret this. A users buys a cheap android device, it does not integrate well with Google services, or becomes obsolete when the OS is not upgraded, so the user buys an iOS device. Here is how Google fixes this. Provide services to the end user. This is in fact how Google became the powerful ad company it is. Way back when, most ad companies did not provide a service, and were quite obnoxious. This meant that many people tried to avoid them. End users turned off cookies and blocked them outright. To counteract this, instead of provided services, the had name like 2o7 that less sophisticated users had trouble deciphering. Google was innovative in that it provided an increasing range of services in exchange for the end user not blocking ad service. What has happened now is that Google is not provided a high level of services. One of their products, Google docs, which cannot be that expensive to service compared to profits, has not been developed. We all know of other products that have been retired. For instance, MS has Skype and Office 365 for only $100 a year that integrates across all products? What is google offering now? Google, like Apple and MS has to develop a more compelling stack, and convince some users that it is worth money. However, as Google is an ad company, and people expect things for free with ads, this is the source of it's profit. Also, Google tends to not be able to hit a price point. This means that actual interesting products, like the glasses and the original google made phone tend to be far beyond the ability of the average user to purchase.

Rock and roll came out the ability to overdrive amplifiers, electrical then electronic.

The crooning typical of music prior to that came about through the ability of microphones to pick up nuances in tone. Prior to this it was just a bunch of guys playing and singing as loudly as they could to try to get the sound recorded on wax.

The last major fight over the structure of music was 30 years ago when everyone was fighting over the right to sample. This, by and large, was due to the fact that for the first time we had a large archive of high quality recording, and the the technology to mix old and new to create a significantly different product.

I suspect that this revolution will be similar. The structure of music has changed. It has gone from an album format, in which most consumers buy and listen to a compilations of songs, to an a la carte format where listeners buy, or more often just stream, a selected song. This has minimized the importance of creating a cohesive album. While every album had one or two radio songs released as singles, most artists tried to make it part of a whole.

In the future I think software will make it possible to string parts of songs together to make something like a dance mix. Some radio DJS used to do this before it was all computer controlled. So like the album losing it status as the definitive unit, the song will also be a legacy concept, artist getting paid royalties only if a part of the song can be structured to fit in a longer musical composition.

Constructivist, free form, hands on education. It works for many students. This may have relevance because there are not some many things to take apart anymore. We are not on farms where things needs to fixed and children can observe, help, participate, then do it themselves. Hell, even cookies are bought prefab, at most you have to cut them. Kids do not see that if procedures are not followed, the cookies are not good. Even making a loaf of bread would benefit them. Even when I was a kid, you still had things you could solder and actually build, not just plug and play. That said, specific teaching methods for specific students is not the silver bullet for the making sure we pick out the students who are going to be tomorrows tech leader. Unless you are being very selective in the kids to get the top 1% motor skills of anyone under 10 is limited and they are not going to have a great deal of motor skills and the abstractions skills are going to be very limited. We see this in spelling bees. These involve a lot of memorization and a limited amount of abstraction. There is no cause and effect because applying the rules incorrectly does not guarantee failure. But it is an age appropriate way to predict future ability to accomplish high paid simple tasks. Likewise a Rube Goldberg machine is a great way to teach cause and effect to older kids, but again it is concrete. Because concrete is the where the kids are at. Development varies, that is why some kids can learn algebra at 10, and some can't even deal with it at 20, but when one is teaching algebra one starts with hands on concrete, and use the subject to move the student to a more abstract view. So, yes, if we are talking top students, this is a viable method to bring kids up to very high expectations by the time they are 13, but I think it might lack pedagogical validity. Like focusing on the ability to pass a multiple guess reading test. Resourceful kids will complete the task without ever learning anything.

SSD is cheap. This is not an Apple. I just bought a stock SSD 512GB drive for 125. This would tend to indicate that once again Dell is building something that has nice looking specs but is going to have unseen bottlenecks.

Also think about Pascal. It too changed programming as it let a new generation of not very good coders generate applications. Who does Pascal now?

Not sure if Java is better or worse than Pascal. A similarity is that part of it's popularity is that it is a teaching language, perhaps more than a production language.

The point of the article is not the goal, but how successful we will be in the short term. However, the goal is an issue in terms of fiction and reality.

First, funding for basic research will continue. The military is probably the most fiscally irresponsible entities on the planet, and so labs claim military application for snail poop. The research money is not significant compared to the entire budget, and some of these things might work, so funding continues and labs continue to do work that will produce nice technology, but for the most part not useful to the military.

Second, the military just everyone else is obsessed with the science fiction possibilities of lasers. This, however, brings up what the real application of lasers are, and why they are used in space opera instead of projectiles. Lasers don't have recoil, so precious propellent is not wasted to provide opposing momentum. Laser cannot be seen before they reach their target, so there is no option for countermeasures prior to being hit. The mean free path in space, even in LEO, is measured in kilometers, light years for deep space, so dispersion is not a problem. Spacecraft can be expected to travel along a predictable path, so one can expect 100 milliseconds to hit the same spot and do damage.

So lasers are useful in space, and the only fiction that exists is the magical power source that provides the endless lasing in a ship no bigger than a caddy or a gun no bigger than an old brick cell phone.

But what are the benefits of more terrestrial warfare. Pretty much nothing near the surface of the earth. The one place it makes sense is in the arena of anti-ICBM defense, but only when we consider the possibility of a nation with a few ICBMs, and only able to launch one at a time. In such a scenario a single launch will be detected, confirmed, and tracked within about 300 seconds, not impossible. Once tracked, a fleet of high altitude laser house in large jets will target the laser and fire within 100 seconds. The speed of the lasers is important because once the ICBM ends boost phase and deploys the warheads, including decoys, it would be difficult to prevent collateral damage from the destroyed bits, if the warheads could be destroyed at all.

Even this realistic application is not yet feasible, and it's limited scope may make it unreasonable. We are taking 10s of billions of dollars to defend against North Korea.

Some uniform rules are good, but many are used to simply exclude a student from an education. In a comprehensive education system, it is a complex dance between the school, the student, and the parents. Compromises are made to maximize education and meet the expectation of the parent. If the parent think it is important for their student to have pretty shoes, that is very difficult to overcome.

Likewise, phone use is complex. I am much more concerned with the student who never turns off the phone at home than a student who uses the phone at school. Phones should be turned off between something like 10PM and 7AM so students can sleep. We have seen suicide cases in which the student simply would not turn off the phone. Parents have to model and enforce this.

As far as phones at schools, they need to teach proper phone use. We cannot turn time back. Students are going to have access to phones and computers at college and at work. if they do not have successful strategies to manage their time and use these tools wisely, they will be less likely to succeed. The loss of five days of class time is miniscule compared to the loss of a scholarship or tuition for a semester.

15 years ago many tech workers screamed the libertarian anti-union songs about how a new economy, not based on old rules like profit and loss and robber barons, would create a magical world in which workers and bosses were equal, and all would be fairly treated, and no governement intervention would be required. Then the bubble burst and people lost jobs and workers had to pay huge taxes on stock incentives that were now worthless and everyone started crying about H1B taking their jobs, just like the lowly auto manufacturer workers. It became a time where there was less difference between a tech worker and bluer collar worker. They were all semi-skilled workers

Here is the thing. Apple, Facebook, MS, are all employers just like any other employer. They want to acquire employees at the least cost. They want to pay the least they can. they don't want people to leave. If this can happen with local employees, that is great. They are cheaper to acquire. But local employees know how much it costs in the US and can leave at any time. That means they cost most in the long term. It would be one thing if local employees could be contract, but the courts have said they can't if they don't have control over the schedule. It would be one thing if local employees could be tied to a job, but courts has awarded money for anti-poaching schemes.

So what is an employer to do. H1B is a good solution. Workers don't know how much the cost of living is, and is likely to be willing to live a much lower standard of living for a certain amount of work. Workers are much harder to poach. Workers are much less likely to complain about an employer violating the laws of the US.

So no, it is not wrong for these companies to want H1B employees. There are not enough US kids who are willing to do a days work that also have mad technical skills. And no, it is not wrong to encourage US kids to go to school and learn the latest technical skills. Even if they do not use them directly, and really many college graduates don't work in their field of study, these skills are useful not matter what. It is also wrong to live in a country where we think that workers do not have a right and need to organize in cartels just like employers do.

MS Windows, by definition, has to run acceptably on crappy hardware. That is the selling point. You can buy a laptop made with reject parts for $300 and it will run. Therefore that is where MS puts it resources. Anecdotal evidence. I ran Windows 7 with Autodesk software in Parallel or Virtual Box. Windows 7 ran very well, I don't know if ran better than Mac OS X, but maybe it did. Autodesk stuff rendered fast, and had very little lag in modeling. Much better than most other laptops. But most other people did not have the speed, memory, or dedicated graphics card. For what I do, the Mac runs better overall than when I try to use my Windows laptop. The only thing the Mac really seems to suck at is transferring large amounts of data through USB. I wish Apple would fix that problem. It has been well known for several years that the best machine for MS Windows is running virtually on a Mac. If the MS Windows image gets borked,which it regularly does because of viruses, it is easy to copy the backup that was made an hour ago by Time Machine back and get to work.

The rational for AA for universities is that they have a mission to educate the leaders of tommorow for the United States and the world. Therefore every real university that I know of tries to bring in a diverse mix of cultures. That is why the labs that I learned in has Asians, whites, American black, Hispanics, Russians,Algerians, Nigerians, etc. I learned not only my hard science and computers, but also how a wide range of people lived, often in significantly different ways from myself.

Some people do not agree with this. Some people think that are are objective instruments in which we can measure certain attributes that can accurately rank applicants. We should therefore only use those instruments to choose the top candidates.

However, most people who work with such assessments knows that this is a myth. There is no objective measurement be it an SAT score, and ACT score, GPA, or IQ test. Therefore universities use a wide range of subjective measure to create a broader ranking. One of the subjective measures that are used is the cultural diversity that the student will bring to campus. Saying that this is not valid is like saying giving credit for volunteer hours, or the essay, or any of the other subjective measures universities use to rank students is invalid.

Clearly these products cannot be sold at a profit. That is why Keurig has to put DRM into the makers. What is heartening is that the number of real suckers are so few that one Keuring imposed the DRM, no one wanted their product. It no longer provided value with respect to competitive products. A lesson we can actually learn, as we have learned from printers, is that if your business depends on selling consumable with a high markup, what you are actually do is creating a vigorous third party marketplace that you are not going to profit. The automatic coffee maker on kickstater was evidently an elaborate hoax that profited only the founders. Sales of Sodastream has plummeted this year in the face of insanely high consumable costs. We will see how the cheap 3D printer manufacturers do and home end users actually figure out how much it costs to print a little widget.

Streaming services are not radio. They are going to replace physical media or even digital media files for the new generation. These services are not cheap advertising, so there is no way that payments to copyright holders are going to be minimized. We are going to need compulsory licensing and reasonable payments to right holders. The lack of such is what is causing the friction and inefficiency, and what is causing loses. I think the argument of whether streaming services are the future is over. This is what kids are used to, and what they will lobby for when they are older.

That said, small payments don't work. The administrative costs of any payments are huge, and customers rightly demand a expensive level of service once payments are made, even if those payments are very small. Probably a $15-$20 a year would cover some costs, but would still require ads. What might be more workable is a sponsored product. A branded credit card would do it.

Yearly payments could work. People pay for sub-basic cable that only supplies local channels. People pay for Hulu. But Hulu plus is almost $100 a year.

This is what seemed to happen to a few other startups, you know the ones that had lists of bars and restaurants and the like, and then tracked attendance and such. The big problem with them was that many of the smaller venues were widely inaccurate, like to the point of putting a bar in someones house. This happened quite a bit in my old neighborhood and I would send in corrections. I think at one point I sent in about 10. Nothing ever happened to fix the problem. I guess they had to staff to make information accurate. The apps were for entertainment purposes, and now at least on these entertainment apps more. There was a time when I wished I treated google maps as entertainment only, and usually I take anything on it with a large grain of salt.

MS is known to break MS Windows every few years, then taking a few years to fix it. We See this with MS Windows 8, MS Vista, Windows ME, Windows 98. This is OK Because most commercial users stick with a version that works. Historical stochastics predict that MS Windows 10 will be a workable version, around 2017, but I expect to be on MS Windows 7 until then. My understanding is that MS Windows X is going to be a rolling upgrade. I interpret this to mean that when one uses the update service, new features will be included with bug fixes and the like. The will be no way to avoid a product like Vista and work will come to a halt. This is different from Mac OS X, where each version is a separate, if now free and almost forcible pushed on users. However, it is not hard to stay with the old version for the lifetime of a machine.