(It really is too bad Cyan didn't do a 1440x1080 rerender / "HD remaster" of Riven. Yeah, it'd have to be 4/3 since they can't redo all their shots, but if they have the art assets it would be a big boon to have an edition with over 2x the vertical resolution.)
Uber is showing, how the consumer feedback, that's easy to provide and is immediately available to anyone with a smart phone,
Right- anyone. That's exactly the problem. All you need to do to game the system as an Uber driver is put together a network of colluders to give you good reviews after you give them "rides". In the past, you only needed to find a few bad actors within the government- now literally anyone can help you with your racket.
Attitudes towards correspondence courses don't change. News at 5.
For the record, correspondence courses have been around since 1892. But somehow MOOCs are "disruptive" (have classrooms and disruption ever gone well together?). Here's a quotation from Wikipedia to add context:
In the United States William Rainey Harper, first president of the University of Chicago, developed the concept of extended education, whereby the research university had satellite colleges of education in the wider community. In 1892 he also encouraged the concept of correspondence school courses to further promote education, an idea that was put into practice by Columbia University.[12][13] Enrollment in the largest private for-profit school based in Scranton, Pennsylvania, the International Correspondence Schools grew explosively in the 1890s. Originally founded in 1888 to provide training for immigrant coal miners aiming to become state mine inspectors or foremen, it enrolled 2500 new students in 1894 and matriculated 72,000 new students in 1895. By 1906 total enrollments reached 900,000. The growth was due to sending out complete textbooks instead of single lessons, and the use of 1200 aggressive in-person salesmen.[14][15] There was a stark contrast in pedagogy:
The regular technical school or college aims to educate a man broadly; our aim, on the contrary, is to educate him only along some particular line. The college demands that a student shall have certain educational qualifications to enter it, and that all students study for approximately the same length of time, and when they have finished their courses they are supposed to be qualified to enter any one of a number of branches in some particular profession. We, on the contrary, are aiming to make our courses fit the particular needs of the student who takes them
Most notably psychology, economics, mathematics and beer brewing. In fact, most of the developments in stats have come about as a result of a need arising in a different discipline. Stats is inherently an applied discipline, so this is not unusual.
What is concerning is how many statistical tools, each with their own set of assumptions, have blossomed up within the past few decades. There are so many stats now that stats can no longer be an ancillary to other disciplines- it needs to be given its own space and statisticians need to be given respect for their unique expertise. There is simply too much knowledge in that domain for those in more theory-driven fields to be able to claim both expertise in the conceptual models of their fields and statistics.
Considering how small the population size for machine learning researchers in academia can be, it is very likely that anecdotes can constitute a satisfactory sample.
One of the most important things you can possibly learn is how to tell when your source is wrong, (optionally) call them out on it, and find another source. If you can't do that, you'll forever rely on others to tell you what should be important to you and spoon-feed you "facts" about those things.
So what you're saying is that if I don't like the bullshit this idiot is spoonfeeding me, I should start listening to the bullshit some other idiot wants to spoonfeed me. Your position is logically inconsistent from my perspective- you're always going to rely on others to tell you what's important and give you information (unless you're a solipsist). I find it interesting too that you think that information sources are either categorically wrong or right- it's really way more fuzzy than that, and ultimately one's ability to determine the signal-to-noise ratio of a source is what allows one to glean the juicy bits.
Many colleges have on-campus currencies- usually dollars are attached to a student ID card via Blackboard or a similar piece of software. I don't use campus currency from my alma mater in my present day-to-day life. How would this be any different?
I heard that you lived in Ithaca at one point in your life. As you might know then, it can be quite snowy in Ithaca.
Do you have reason to believe that superstructures (in the style of Buckminster Fuller) could be built using 3D printing in the future, or are the technical challenges too great? Could a dome be printed? Is there any hope for keeping out at least some of the snow (perhaps not all of it as it is sometimes quite nice)?
On second thought, I just made a pretty good argument for more training in domain-specific languages. I still don't think that just plunging in like a mid-90s cowboy coder generalist will work anymore though. Those days are definitely gone.
I think that you're making the mistake of perceiving a mindless code monkey to be tantamount to someone who is a seasoned computer scientist with a solid grasp of theory and a fair understanding of software engineering principles/design patterns (or a super competent software engineer with a fair understanding of theory). Code monkeys will not make real discoveries or do real work- like it or not, for better or worse, only the super-talented will (yeah, reality's a bitch). We've also reached a bit of a ceiling effect in science and tech more generally in my eyes- all the low-hanging fruit has already been picked, so the discoveries that remain to be made require much more effort and interdisciplinary teamwork than ever before. Getting more people trained to code won't change that.
The other point is that most programming languages these days are becoming more expressive anyways, which lowers the entry barrier to coding significantly so that most people will be able to figure it out at one point or another anyways- you don't need to be in the IQ > 120 club anymore because you don't need to really understand pointers or assembly code or any of that mess. Domain-specific languages are becoming mature enough that a statistician won't necessarily need to learn C and can most of his or her work done with R; ditto for the scientist who wants to use Julia or SciPy (without delving into any of the non-SciPy libraries available in Python). Syntactic sugar has been added to web languages like such as Javascript (e.g., Coffeescript) and even HTML/CSS (although goodness knows why these needed syntactic sugar). Perhaps I'm just coming from a privileged standpoint where I already find it simple so I can't see how other people will continue to find it hard, but I really really don't think that the simpler aspects of programming are going to be out of reach for the masses that much longer.
One last point is that a lot of the progress I've noticed in the tech world right now seems to be in the world of DevOps, which is what I believe is being referred to in point 2; a minimal number of systems administrators and developers are needed now to due to advances in deployment and debugging automation. Case in point: Google's servers broke and fixed themselves. Do we still need workers to do these tasks now? Definitely. 10 years from now? Not so sure, and flooding the job market with a bunch of "coders" certainly won't make matters better.
Care to explain how this is true? I think I have an idea, but using "a healthy combination of certain areas of comp-sci (databases, machine learning, NLP, AI), statistical methods, and, quite often, improvisation" seems to be an even more obtuse approach than going about it the old-fashioned way. I'd much rather hear that people are using what we already know or (still better, but probably not as plausible) the latest mathematical advances regarding nonlinear systems rather than just ad-hoc'ing methods because... computers! I believe that this is at least partiallly Nassim Taleb's objection to the entire field of data science as well. How many 'results' coming from data science are the product of sound and rigorous methodologies, and how many are just due to chance/data dredging?
Slashdot Mirror
I know what the next generation needs too!
https://www.youtube.com/watch?...
(It really is too bad Cyan didn't do a 1440x1080 rerender / "HD remaster" of Riven. Yeah, it'd have to be 4/3 since they can't redo all their shots, but if they have the art assets it would be a big boon to have an edition with over 2x the vertical resolution.)
My friend, I give you the Starry Expanse
...actually hire and pay editors like a normal encyclopedia rather than focus on improving an already mature enough web application.
Sorry, I think I have a case of the Fridays.
Stop making me want to gouge my eyes out when I look at billboards. That is all.
It's one of the same problems that Diaspora encountered. Widespread adoption is a bitch.
See: http://abandonmatlab.wordpress.com/
"Data Scientist" being, of course, a rebranding of the term "Data Analyst".
The witty banter and mercurial charm?
Uber is showing, how the consumer feedback, that's easy to provide and is immediately available to anyone with a smart phone,
Right- anyone. That's exactly the problem. All you need to do to game the system as an Uber driver is put together a network of colluders to give you good reviews after you give them "rides". In the past, you only needed to find a few bad actors within the government- now literally anyone can help you with your racket.
For the record, correspondence courses have been around since 1892. But somehow MOOCs are "disruptive" (have classrooms and disruption ever gone well together?). Here's a quotation from Wikipedia to add context:
In the United States William Rainey Harper, first president of the University of Chicago, developed the concept of extended education, whereby the research university had satellite colleges of education in the wider community. In 1892 he also encouraged the concept of correspondence school courses to further promote education, an idea that was put into practice by Columbia University.[12][13] Enrollment in the largest private for-profit school based in Scranton, Pennsylvania, the International Correspondence Schools grew explosively in the 1890s. Originally founded in 1888 to provide training for immigrant coal miners aiming to become state mine inspectors or foremen, it enrolled 2500 new students in 1894 and matriculated 72,000 new students in 1895. By 1906 total enrollments reached 900,000. The growth was due to sending out complete textbooks instead of single lessons, and the use of 1200 aggressive in-person salesmen.[14][15] There was a stark contrast in pedagogy:
The regular technical school or college aims to educate a man broadly; our aim, on the contrary, is to educate him only along some particular line. The college demands that a student shall have certain educational qualifications to enter it, and that all students study for approximately the same length of time, and when they have finished their courses they are supposed to be qualified to enter any one of a number of branches in some particular profession. We, on the contrary, are aiming to make our courses fit the particular needs of the student who takes them
http://en.wikipedia.org/wiki/D...
Most notably psychology, economics, mathematics and beer brewing. In fact, most of the developments in stats have come about as a result of a need arising in a different discipline. Stats is inherently an applied discipline, so this is not unusual.
What is concerning is how many statistical tools, each with their own set of assumptions, have blossomed up within the past few decades. There are so many stats now that stats can no longer be an ancillary to other disciplines- it needs to be given its own space and statisticians need to be given respect for their unique expertise. There is simply too much knowledge in that domain for those in more theory-driven fields to be able to claim both expertise in the conceptual models of their fields and statistics.
Considering how small the population size for machine learning researchers in academia can be, it is very likely that anecdotes can constitute a satisfactory sample.
One of the most important things you can possibly learn is how to tell when your source is wrong, (optionally) call them out on it, and find another source. If you can't do that, you'll forever rely on others to tell you what should be important to you and spoon-feed you "facts" about those things.
So what you're saying is that if I don't like the bullshit this idiot is spoonfeeding me, I should start listening to the bullshit some other idiot wants to spoonfeed me. Your position is logically inconsistent from my perspective- you're always going to rely on others to tell you what's important and give you information (unless you're a solipsist). I find it interesting too that you think that information sources are either categorically wrong or right- it's really way more fuzzy than that, and ultimately one's ability to determine the signal-to-noise ratio of a source is what allows one to glean the juicy bits.
Many colleges have on-campus currencies- usually dollars are attached to a student ID card via Blackboard or a similar piece of software. I don't use campus currency from my alma mater in my present day-to-day life. How would this be any different?
This too:
http://www.ted.com/talks/lawre...
In an ideal world, yes.
In the interim, I like what Larry Lessig has to say.
I heard that you lived in Ithaca at one point in your life. As you might know then, it can be quite snowy in Ithaca. Do you have reason to believe that superstructures (in the style of Buckminster Fuller) could be built using 3D printing in the future, or are the technical challenges too great? Could a dome be printed? Is there any hope for keeping out at least some of the snow (perhaps not all of it as it is sometimes quite nice)?
If you're asking Slashdot to make major life decisions for you, you have bigger issues. That said, why the hell not.
Wow. What was I thinking this morning? These opinions are crap.
On second thought, I just made a pretty good argument for more training in domain-specific languages. I still don't think that just plunging in like a mid-90s cowboy coder generalist will work anymore though. Those days are definitely gone.
I think that you're making the mistake of perceiving a mindless code monkey to be tantamount to someone who is a seasoned computer scientist with a solid grasp of theory and a fair understanding of software engineering principles/design patterns (or a super competent software engineer with a fair understanding of theory). Code monkeys will not make real discoveries or do real work- like it or not, for better or worse, only the super-talented will (yeah, reality's a bitch). We've also reached a bit of a ceiling effect in science and tech more generally in my eyes- all the low-hanging fruit has already been picked, so the discoveries that remain to be made require much more effort and interdisciplinary teamwork than ever before. Getting more people trained to code won't change that.
The other point is that most programming languages these days are becoming more expressive anyways, which lowers the entry barrier to coding significantly so that most people will be able to figure it out at one point or another anyways- you don't need to be in the IQ > 120 club anymore because you don't need to really understand pointers or assembly code or any of that mess. Domain-specific languages are becoming mature enough that a statistician won't necessarily need to learn C and can most of his or her work done with R; ditto for the scientist who wants to use Julia or SciPy (without delving into any of the non-SciPy libraries available in Python). Syntactic sugar has been added to web languages like such as Javascript (e.g., Coffeescript) and even HTML/CSS (although goodness knows why these needed syntactic sugar). Perhaps I'm just coming from a privileged standpoint where I already find it simple so I can't see how other people will continue to find it hard, but I really really don't think that the simpler aspects of programming are going to be out of reach for the masses that much longer.
One last point is that a lot of the progress I've noticed in the tech world right now seems to be in the world of DevOps, which is what I believe is being referred to in point 2; a minimal number of systems administrators and developers are needed now to due to advances in deployment and debugging automation. Case in point: Google's servers broke and fixed themselves. Do we still need workers to do these tasks now? Definitely. 10 years from now? Not so sure, and flooding the job market with a bunch of "coders" certainly won't make matters better.
No
Perkele.
not amenable to analysis using classical methods.
Care to explain how this is true? I think I have an idea, but using "a healthy combination of certain areas of comp-sci (databases, machine learning, NLP, AI), statistical methods, and, quite often, improvisation" seems to be an even more obtuse approach than going about it the old-fashioned way. I'd much rather hear that people are using what we already know or (still better, but probably not as plausible) the latest mathematical advances regarding nonlinear systems rather than just ad-hoc'ing methods because... computers! I believe that this is at least partiallly Nassim Taleb's objection to the entire field of data science as well. How many 'results' coming from data science are the product of sound and rigorous methodologies, and how many are just due to chance/data dredging?