rsanders

If you read the recent Time cover story about medical costs, you'll see examples of people being recursively double-billed for items, at least according to the Medicare guidelines. e.g., and my memory's a little fuzzy on the specifics, but I believe one case was being billed for the operating room, then billed for the prep kit (which is supposed to be included in the room charge), then billed for every individual item in the prep kit.

If I hire a contractor to do something and he charges $X/hour, I don't expect him to mark up the materials he might use, too. I'm already paying for his time and expertise. I shouldn't also be paying the drywall for his expertise.

Is anybody using this for Clojure? I'm very interested in how usable it is for that.

Clojure runs on the JVM using JVM bytecodes, but that doesn't mean Takipi will be able to show me anything but mangled gibberish.

I attended GT a long time ago (early 90's), so I might be out of date, but I wouldn't go far as to call it "culture". I kid, I kid! :-)

It's a notoriously demanding program. I hope they can adapt their "Take a look at the people to your left and right. One of you will not make it to graduation." scare speech/boast to the online format.

There's nothing I could learn from the Coursera compilers class I couldn't have learned from a $10 used textbook, and yet the Coursera class did make a huge difference. As Good Will Hunting sort of mis-stated it -- you can get access to a $150,000 education's worth of information for the cost of a library card and some late charges. Or, given Khan Academy and all the MOOC companies, with a cheap netbook and an Internet connection. None of that information may cross the book/brain barrier, of course.

This clearly isn't a completely replacement for the existing Masters program. Yet it's a huge improvement over a stack of dusty books, and possibly, in some cases, an improvement over a stack of "Certificates of Accomplishment" from non-accredited MOOCs.

I don't think this is necessarily aimed at people just exiting a BS program. It has a strong "professional" focus. If you haven't learned to interact with people fruitfully and work in teams in your professional life, you're at a disadvantage.

As I said elsewhere, I believe a program like this requires more of the student than a traditional degree. You have to bring greater discipline, the ability to learn without interaction with the instructor, and as you point out, an assortment of other skills that you have acquired on your own.

I do hope that online educators find some way to foster similarly valuable interaction between students on team projects as opposed to keeping everybody in individual solipsistic relationships with the course automaton. It's a challenging problem -- who wants to depend on somebody without any, or much, skin in the game? How would it even work with the flexible schedules this kind of thing will attract?

That appears to be the intent. See my other comment: https://news.ycombinator.com/item?id=5709930

I fit into that category, and I'm taking 1-2 significant MOOCs (compilers, ML, data science, etc.) at a time right now. I can just make it work, but I wouldn't if I had to show up for class. The fixed schedule format of Coursera adds a bit of pressure, but probably also a bit of motivation.

I agree that online classes from GT aren't guaranteed to be better than online classes from other places. Perhaps one day we'll see an a la carte degree offered by Udacity, so you can pick the courses from the universities or instructors you prefer.

That said, branding and standardization are important. Even if the classes are no better, the perceived value of the degree might be higher than one from Podunk State Online.

It's described here: http://www.cc.gatech.edu/future/masters/mscs/program

The OMS won't initially have all the same courses (or specializations?), but the plan according to one of the GT videos is to reach approximate parity over time.

Georgia Tech already has a distance learning MS in CSE, but not a general MS CS. And I believe the cost is rather higher than this.

http://www.cseprograms.gatech.edu/distancelearning

First off, the MOOCs associated with this degree will still be free. The degree is not. The change here is that a) an entire MS worth of course material (and the systems to support it) from a respected institution are being created, b) the actual (O)MS degree just became a great deal less expensive.

It remains to be seen whether this lowers the bar or merely opens up more opportunities. I don't think education is a zero sum game. And while there may be a glut of PhDs, there is a shortage of employees with the advanced knowledge and skills in demand now. That's why e.g. AT&T is pushing this program.

While the union issue and age discrimination are interesting and important topics, CS graduates don't have a lot to complain about economically. The field has maintained high employment and excellent compensation during recent times when many other professions suffered greatly. Now, at 40 I don't expect to get funded by Y Combinator because I've almost certainly suffered crippling age-related innovation dementia, but there are plenty of employers that don't think that way.

There have been several articles posted here about the glut of PhDs versus available faculty positions. And actual instruction is a pretty small part of what full professors are expected to do. If you aren't pulling in research grants, your career is on a weak trajectory. So the problem was already there, and I don't think this seriously affects it one way or the other.

One might hope that making education available and practical to many more people at much lower costs might actually help advance the field, both within and without the ivory tower.

You're right -- that was phrased ambiguously. I should have said "there will be an admissions process similar to that for the non-online MS CS". I don't know what the exact requirements will be.

This degree program doesn't allow people without a BS to get a graduate degree. I think it's wrong to say that having an entire Masters degree program from a highly respected institution online and accessible to everybody, everywhere for free doesn't help autodidacts.

The biggest impact for those who don't qualify for admission may be in advancing the legitimacy of online education, with or without a degree.

As an employer, the difference between an awarded MS and a certificate of completion in this program is not going to be that substantial to me. I'm going to interview both candidates to determine their level of knowledge and skill, and both indicators communicate the same educational background.

I expect a natural outgrowth of this sort of thing will be students providing access to his/her detailed course records as part of their resume. If I'm interested in hiring developers for machine learning applications, being able to read the code from those standardized, well-known assignments would be tremendously more useful in skill assessment than asking somebody to code FizzBuzz on the whiteboard and "riddle me this, batman: why are manhole covers round?"

The "GitHub account as resume" idea is an interesting trend, but they're not standardized and I don't have time to digest 20 unique projects in 20 different languages on 20 different frameworks to ferret out the non-boilerplate bits which may tell me something about the candidate.

From the Inside Higher Ed article I linked elsewhere:

> Galil and Bras both said they did not think the Udacity partnership would cannibalize its existing master’s degree program. They said the residential college experience would still appeal to students who are on a Ph.D. track, who want a full range of career services and who want visas to live in the United States while they study.

Apparently the GRE will not be a requirement. From http://www.insidehighered.com/news/2013/05/14/georgia-tech-a...:

> The second type of student will be “prospective degree-seeking” students who will be admitted to the program tentatively because they will not have to take the GRE as other applicants do. If they do well in two core classes, Georgia Tech will put them on the degree track. The university expects to enroll 2,000 such students in the next three years.

According to http://www.insidehighered.com/news/2013/05/14/georgia-tech-a...:

> Georgia Tech said it does not plan to lower admission standards to find 6,000 or so students for this track -- a number than is 20 times larger than its current computer science master’s degree program. Instead, Georgia Tech hopes to attract more qualified applicants from across the world, including inside the military and at companies – places that harbor nontraditional students who could not previously come to a traditional campus or find the money for a full degree, on campus or online.

I wish I'd had this article to post instead. It has some detailed information about the cost of a traditional MS in CS (the online version is "one sixth the cost"), revenue share with Udacity, admission requirements, and a lot more.

  http://www.insidehighered.com/news/2013/05/14/georgia-tech-and-udacity-roll-out-massive-new-low-cost-degree-program

I agree -- but we can't really know about that until the program's run for a while.

I think they made a pretty smart choice starting with a CS degree. The field is notoriously full of extremely talented autodidacts, whether because the material is more amenable to that approach or because those who learn it are more capable of doing so.

In any case, the attitude towards educational background in the field of CS/software engineering is already quite liberal and accepting. So there's less of a prejudice against this program than there would be for an online MS in many other fields. Now they just have to make it work.

Despite other comments about not staking their reputation on this degree -- and I agree that the "Online" part of the title seems like a hedge -- Georgia Tech takes its reputation extremely seriously. Smart people are staking their own personal reputations and careers on this program. This is something new and it may fail, but not because there is any intent to deliver a lesser education.

I can ask about that (my wife works at GT PE), but this video implies a pretty strong equivalence. The relevant part is right at the beginning.

http://www.youtube.com/watch?v=zihmiLx3Xlk

The course materials and participation are MO, just not the Masters degree itself.

At the next family-attending Georgia Tech Professional Education function I'm going to petition everybody who'll stand still to allow people without a BS to earn the degree. But I think I'll actually be preaching to the choir.

Anyway, it's a huge step for them to create this program as is. And once the concerns about "degree devaluation" have passed, and enough demand has become apparent, maybe they'll be ready for the next step.

Well, there's no real scarcity of teaching faculty and other resources in a MOOC format, so that pressure will be removed. But I don't know how the standards will be set. I'm not sure Georgia Tech knows that yet, either.

Yeah, I find that a little wishy-washy, too. I guess that's part of the buying decision you have to make -- are you willing to risk the lesser cachet for the convenience and affordability?

I'd do it in a second, but then I'm probably the target audience (other than not having completed my BS). I'm almost 40 years old, with a wife and child, and the flexibility is crucial to me. This isn't an alternative to a traditional MS program; the traditional MS just isn't even an option for me. I see it as a way to get credit for the MOOC work I'm already doing for fun and fulfillment.

Unfortunately, at least for the moment an undergrad degree is a prerequisite. More generally, you have to be admitted to Georgia Tech as would any other graduate student.

I think there will be quite a few companies willing to pay $7000 in tuition assistance so their employees can get a MS.

The word "devalue" comes up a lot in conversations about this sort of thing, but I'm not sure exactly how that works. If half the student body at GT died before graduating, would your degree be worth twice as much?

I think the Masters degree is an excellent place to start. MOOC-style education requires more of the student than traditional teaching formats. A lot of people don't have the necessary skills leaving high school, but they are likely to have developed them during the course of an undergraduate degree.

I think the real reason you have AT&T involved is that without that sort of influence, this degree program wouldn't have happened until years later. Universities, and boards of regents, are deeply conservative, risk-averse institutions in some ways. There's also a lot of cost to startup a program like this, and public universities aren't exactly rolling in money right now.

Beyond that, a lot of companies have some kind of tuition assistance for graduate degrees, and a lot of companies are having trouble finding skilled talent to begin with. So it's a way to save money and improve recruiting.

The business models of companies like Udacity and Coursera are probably going to rely heavily on the hunger of companies the size of AT&T, Google, etc. for a steady supply of trained workers. Somebody's got to pay the bills and make the VCs rich.

I like to think of this as shifting part of the cost of the education from the individual to the company that's ultimately benefitting from it. Of course, one might worry about some unspecified negative influence from the corporate sponsors, but in this incredibly transparent educational environment, such influences would be pretty obvious and self-defeating.

The cost of a similar master's degree at a university like Georgia Tech can run $40,000.

It's not quite the same thing -- watch the video by Charles Isbell on the GT site to hear their differentiation between the "on campus" and online degrees. But it's an amazing value for people who can learn effectively in a MOOC environment and aren't focused on the path into research and academia.

The video is also at http://www.youtube.com/watch?v=zihmiLx3Xlk.

[edit: added link to video]

If you chase the links you end up on the Georgia Tech site about it: http://www.omscs.gatech.edu/faq.

  We’re not yet ready to announce a specific program cost, 
  but the plan is to offer the Georgia Tech OMS CS for a 
  total cost of under $7,000—a fraction of the cost of 
  Georgia Tech’s on-campus program and even less than that 
  of comparable private universities.

All the content will be free to everybody, but you have to apply to Georgia Tech and pay if you want the actual degree.

You also have to have a Bachelor's degree, which is a disappointment for me. I left college before completing my BS in CS, and I'd love to be in this program, but I wouldn't be admitted.

[disclosure: my wife works in the Georgia Tech Professional Education group, which is how I heard about this]

In the original post, he said "we could ship...". That's a pretty clear disclosure of his Googleness to me.

So you never see conflict messages or conflicted files in your Dropbox? In the few cases I've tried a git working copy on Dropbox -- which is what I believe the author is discussing -- I've had conflicts.

For example:

    $ ls -1 .git/index*

    index
    index (blacktip.esquimaux.org's conflicted copy 2013-04-09)
    index (hammerhead's conflicted copy 2012-01-30)
    index (holodeck's conflicted copy 2012-01-26 (1))
    index (holodeck's conflicted copy 2012-01-26)
    index (holodeck's conflicted copy 2013-04-09)

It was only the index file and not anything irreplaceable (not that git does a lot of mutating of files in general). But it rapidly ended my experiment with git on Dropbox.

So you're saying that computer scientists are inherently more likely to be correct about an economic proposition than economists? Could it be that the field of computer science is more amenable to provably correct statements than the field of economics?

I'd say that computer science has shown similar failures when dealing with phenomena that are as large-scale and chaotic as economics. Artificial intelligence is one such area in which it has been difficult to produce results.

Using the word "system" as broadly as you have in this post seems like a recipe for shaky reasoning.

It appeared to be syncing my entire Cloud Player music folder to my local machine. That's quite a few GB of data -- 12,000 songs or so out of a possible 250,000. So it's not going to be of much use to me until they add some kind of selective sync option.