Imagine a Java coder. Took courses exclusively in Java, aimed at knowing every minute subchunk of the API. "Graduates" from his trade school. Knows nothing of the larger art/science/cruft of computer science. Couple of years into the workforce, she wants to quit coding Java.
She's giving up an education. I would submit that it was a el crapola education to begin with, but she's still giving it up.
I happen to believe that GOOD educations stick around through general conceptualizations, rather than rote memorization, but that's gotta be drifting, if not steaming, OT.
What companies don't see immediately is that by sending stuff offshore, you're introducing a raft of problems:
Intellectual Property Theft: Teach non-employees the X algorithm, which gives you most of your competitive advantage, and it'll spread. Since the contractors can move to another firm more quickly than a sizable number of your actual employees, it'll leave even faster. Yes, NDAs, etc., but that'll get tied up in court until it's irrelevant.
Lock-in: What happens when the guys to whom you're outsourcing suck? Even worse, what happens when you've been happy for two years, THEN they start to suck? Switching costs are non-trivial. There was an article on how, of Australian (I think) companies outsourcing their programming, 80% planned to continue, but all of those planned to renegotiate the contract. Growing pains for these relationships exist, and that's a cost, too.
And that's just off the top of my head. These costs don't show up until you leave the outsourcers. Right now, they're saving $n/programmer-hour, but there are more subtle costs they're incurring en route. Just my 2 rupees.
Partly this is the fault of the music industry, yes, for pushing crap. Partly it's the fault of musicians themselves, who are more interested in being Rock Stars(tm) than in making actual music.
What makes you think that the collective output of the bands will be any different under a different revenue stream? Suddenly, Tom Waits becomes idolized by my SO for his musical talent?
Face it:
Musician's security == more sales == mainstream sound
Period. Big sales go to populist artists, niche sales go to niche players. Even worse, with 150,000 people recording, you (personally) cannot hear enough music to know everything that's out there, so marketing becomes even more important than it is now.
All the filesharing revolution does is centralize power in the hands of the marketers. I, at least, cannot find copy 1 of Blood Money on Sharaza, although I'm sure there are a billion copies of "The Ketchup Song."
Distribution policies are irrelevant. Consumer tendencies matter. The latter won't change due to the former.
Sure, there are some bands who seem to buck those trends, but when you're looking at the real longevity of bands like Aerosmith, versus the possible (and tenuous) longevity of artists like Britney Spears... well, you know what? I think in ten or fifteen years, there will still be people listen to old Aerosmith tunes, but Britney Spears will be all but forgotten.
Note, however, that you'd likely have said the same thing about Madonna back in the "Holiday" or (hell, that other song that was a hit off her first album, circa 1986) era. I'm certainly not going to say that she's artistically valid, but you can't deny her longevity.
I've thought about it a bit, and it appears to me that people just flat out get tired of the stars we're talking about, especially the teenyboppers. Just look at what happened to Radiohead. For two YEARS, they were everywhere. Lately, they've disappeared.
Plus, don't ignore the fact that the industry isn't geared towards producing quality music. In country, for example, are you better off promoting Hank Williams, Sr., or Garth Brooks?
Sure, Hank's recordings are already in the can, but Garth can produce totally forgettable country tunes that you can promote the hell out of, sell a bunch of copies of, and then move on to the next shooting star.
Disgusting, but that's what that industry's business model is currently geared for.
I played with this idea briefly. Imagine a high end hotel offering "whatever" on demand for sum of cash $x. Networking to the rooms is a solved problem (see spectravision, etc.). Only question is how to get the content.
Well, that, and selling and servicing it in a scalable fashion to hotels that aren't terribly interested in giving you much of a cut.
Not a bad idea, but you run into trouble with the marketing and the amount of time you need to keep things vs. your affordable drive space. Not to mention the copyright issues the networks will come up with.
I used to be like you, but the format on our NPR station (KUT,Austin, TX) appears to have really lightened up since the WTC came down.
i.e. the second national news segment got replaced by some inane (IMHO) local news bit that doesn't really address ANYthing. As such, I've stopped listening as much.
I'm curious if other NPR listeners have seen similar results elsewhere.
I'm really surprised TiVo's prices haven't dropped, too. I'd think there's gotta be a Laffer curve point somewhere in the $300ish range for box and lifetime. I would think they've gotta be able to make money at that level.
The problem MIGHT be that TiVo licenses its HW manufacture, so Philips/Sony has to make coin on the box itself, while TiVo's really about the SW and programming updates.
Still doesn't solve your issues, though. Maybe it'll happen when TiVo starts getting built into TVs and the like. Frankly, I'm a little amazed that hasn't happened yet. That Panasonic PVR/DVD-RW is just dying for TiVo.
We have two TiVos, one 30hr, one 20hr. We've never had a space problem on the big one, and we've had the 2hr finale of the XFiles on there forever. (We just can't get ourselves to watch it).
We have run into issues on the 20hr, but that's mostly because I like movies and sporting events, so the weekend gobbles up a bunch of programming.
Crime-Warner is giving away a Scientific Atlanta PVR (dual tuners, etc etc) for nearly nothing to customers with higher-end packages. Same guide data as Tivo (often I've noticed the program descriptions from my SA2100 box are word-for-word identical with Tivo), and many Tivo features.
Except that it sucks goatse.cx. Seriously. We have one of those, and two TiVos. The TW box we have (our second, after the first one blew up) has all of the following useful feaures:
reboots spontaneously, even while we're watching a program.
mysteriously turns itself off at random times (like Wednesday afternoon), and doesn't come back up. You can't record anything if it's not powered on.
Let's say you want to watch something that it's currently recording. Like, say, an episode of Farscape that's been running for half an hour, but hasn't finished yet. When you tell the box to start watching the show, it dumps you at the END of the recorded portion (aka LIVE TV).
That, we can fix. Just rewind back to the beginning of the show. Slight hassle, but not horrible. Then, when the RECORDING ends (irrespective of where in the program YOU are), it dumps you back to the live feed. Not horrible for regular programs, but it sure sucks when you accidently see the final score of the basketball game you were watching.
So you're halfway through a show, and you go run an errand. While you're gone, your SO watches something else. When you return to watch the rest of your show, the TW box starts at the BEGINNING, not where you stopped watching.
TiVo has a function where you can record beyond the end of your show. College hoop, for instance, tends to run long, so you can tell it to record an extra half hour to make sure you get the end (and OT, if applicable). The TW has a similar function that you can program, but it doesn't work.
TiVo does all of the above admirably, with a user interface my technology stunted inlaws can use. That $500 never made so much sense, and the TW box is going back when I get a spare moment to do it.
Isn't the issue in this area $/MIPS?
on
New SGI Altix 3000
·
· Score: 3, Interesting
I don't do this for a living, but it seems that $/MIPS is the only benchmark even worth discussing, so shouldn't one be able to put together massive clusters of boxen to do the same thing, only without the SGI price tag?
I hate to admit this, but I did an NPV analysis of the various payment schedules. This was back when monthly was $10, lifetime was $200, and (now unavailable) yearly was $100.
The upshot was that monthly was cheaper until month 17-18, and lifetime was cheaper after that. Yearly never made sense. So, if you expect to keep your TiVo longer than 18 months, which plan to get is pretty clear.
It's been pointed out that I should've included an end value for the lifetime subscription. I didn't, but that would just make the expected usefulness of the monthly plan even shorter.
I haven't re-run the numbers for the new pricing scheme, but I expect they'd be similar.
Not really. Your best bet might be keeping track of the Freevo project, although I've no experience with that, and it doesn't appear to be able to tape at the moment.
> TV listings are hardly worth $5/month...
We own two TiVos, and you're not just paying for the listings, but for the superb user interface. As many other posters have already mentioned:
"Yes, you COULD do everything TiVo does with a homegrown computer, but TiVo comes in a squeaky clean WORKING package, as opposed to cobbling something together from assorted parts lying around your garage."
I've had two SW upgrades during the couple of years we've had the TiVos, and both upgrades have brought features that significantly enhanced the use of my TiVo. Easily worth $10/month, IMHO.
>...some company sifting through my TV viewing habits
TiVo are very open about what they track and what they don't. IIRC, they don't even trace back to the user, but use your information in conjunction with others to create profiles.
At that level, I don't have much of a problem with it.
During the Final Four last year, my beloved Kansas Jayhawks were playing. I came home from a happy hour the Friday night before the final four to find that TiVo had recorded the Final Four practices for all four teams. I didn't even know anybody would be nuts enough to cover that non-event. Needless to say, I was thrilled to see KU and their opponent's practices appear, unbidden, on my TiVo.
For that occasional miracle, I'll take all the Univision soap operas, shopping channel dreck, and Korean news I can delete, and I'll thank TiVo every time for trying.
I seem to recall people having trouble with the speed of the "now showing" window after their upgrade to 140hrs. If that's true, I can't imagine what 700 would do...
I once saw a spec (back in my VB programming days) that had something like four or five parameters in the hungarian notation before you even got to the variable name. something like
intLocalFnnameModulenameX
THAT gets a little absurd, IMHO. Not arguing that intX, or even intLocalX isn't useful, but you can twist yourself around an axle pretty damn quickly with this stuff.
have them "use it like it'll get used" for a while. Note all the problems, fix them. From there, you can create use cases that'll let sobody else do the testing.
It's CRITICAL, IMHO, that the people requesting the application get directly involved with how the front ends should work. If they don't, you're just asking for UI rework pain.
These guys sink millions upon millions of dollars into focus groups and various other forms of research to determine what music is going to be the most popular.
I thought about this a bit, and they don't really have to. Follow me down this road...
ClearChannel has a playlist of exactly X songs on their "pop" stations.
Because they can own (virtually) all the radio stations in a given market, they can break them out as they see, assigning one to pop, one to classic rock, one to hard rock, etc., etc..
Given that, there's no competition for "pop" music.
Given THAT, those X songs in the playlist ARE the top X songs in the pop demographic, practically by definition.
No market research, just raw ownership of the airwaves and, by extension, the markets.
And don't give me 'from a friend who heard from a friend' etc., most likely that chain, however long, started with an ad.
I dunno, did you ever see a TiVo commercial? I think there were something like three, and they all sucked. TiVo's marketing has been horrible to date, IMHO.
Touche. OTOH, that perspective makes life into a big, dull distributed.net problem, and ignores what Kuhn (IIRC) called revolutionary science. So you set the rules to reflect our notion of reality now. You can get the bot to work through all sorts of permutations of the variables you assign it, but it won't come up with anything that's actually original.
If your reality didn't reflect that the earth goes around the sun, for instance, the bot couldn't do anything to disprove this. It would simply take the world it was presented, and maximize its reaction. I guess I'm wondering about what happens regarding breakthroughs that don't follow rulesets.
I'm no rocket scientist, but I can't get past the notion that AI simply takes a goal, gives the algorithmic rules that apply to the world, and lets the algorithm go nuts trying out new stuff.
I'm thinking particularly of a genetic model I saw a few years ago, where the goal was "maximize speed," the ruleset provided physical characteristics of the world (i.e. gravity, friction coefficient of the ground, and so on), and while the results were interesting, I'd have trouble characterizing any of that as thought.
As such, when you set a goal of "reasonable conversation," and provide a ruleset and knowledge base, the machine isn't so much "thinking" as internally contesting two reactions to the ruleset.
Bantha poodoo. The video's going to have to go through a TV, to keep the price down in the markets you discuss. If you'd prefer to say "the number of internet capable machines will double by 2008," you've got a shot, as the internet ready PS2/Xbox/GCube/CableBox is where the battle has been for the last three years anyway, but they sure as heck won't be PCs.
Slashdot Mirror
I think he might be talking about this:
Imagine a Java coder. Took courses exclusively in Java, aimed at knowing every minute subchunk of the API. "Graduates" from his trade school. Knows nothing of the larger art/science/cruft of computer science. Couple of years into the workforce, she wants to quit coding Java.
She's giving up an education. I would submit that it was a el crapola education to begin with, but she's still giving it up.
I happen to believe that GOOD educations stick around through general conceptualizations, rather than rote memorization, but that's gotta be drifting, if not steaming, OT.
And that's just off the top of my head. These costs don't show up until you leave the outsourcers. Right now, they're saving $n/programmer-hour, but there are more subtle costs they're incurring en route. Just my 2 rupees.
Face it:
Musician's security == more sales == mainstream sound
Period. Big sales go to populist artists, niche sales go to niche players. Even worse, with 150,000 people recording, you (personally) cannot hear enough music to know everything that's out there, so marketing becomes even more important than it is now.
All the filesharing revolution does is centralize power in the hands of the marketers. I, at least, cannot find copy 1 of Blood Money on Sharaza, although I'm sure there are a billion copies of "The Ketchup Song."
Distribution policies are irrelevant. Consumer tendencies matter. The latter won't change due to the former.
I played with this idea briefly. Imagine a high end hotel offering "whatever" on demand for sum of cash $x. Networking to the rooms is a solved problem (see spectravision, etc.). Only question is how to get the content.
Well, that, and selling and servicing it in a scalable fashion to hotels that aren't terribly interested in giving you much of a cut.
Not a bad idea, but you run into trouble with the marketing and the amount of time you need to keep things vs. your affordable drive space. Not to mention the copyright issues the networks will come up with.
I used to be like you, but the format on our NPR station (KUT,Austin, TX) appears to have really lightened up since the WTC came down.
i.e. the second national news segment got replaced by some inane (IMHO) local news bit that doesn't really address ANYthing. As such, I've stopped listening as much.
I'm curious if other NPR listeners have seen similar results elsewhere.
It's:
Mom is SOL on linux.
At least mine is. Your mom's an astronaut.
I'm really surprised TiVo's prices haven't dropped, too. I'd think there's gotta be a Laffer curve point somewhere in the $300ish range for box and lifetime. I would think they've gotta be able to make money at that level.
The problem MIGHT be that TiVo licenses its HW manufacture, so Philips/Sony has to make coin on the box itself, while TiVo's really about the SW and programming updates.
Still doesn't solve your issues, though. Maybe it'll happen when TiVo starts getting built into TVs and the like. Frankly, I'm a little amazed that hasn't happened yet. That Panasonic PVR/DVD-RW is just dying for TiVo.
We have two TiVos, one 30hr, one 20hr. We've never had a space problem on the big one, and we've had the 2hr finale of the XFiles on there forever. (We just can't get ourselves to watch it).
We have run into issues on the 20hr, but that's mostly because I like movies and sporting events, so the weekend gobbles up a bunch of programming.
- reboots spontaneously, even while we're watching a program.
- mysteriously turns itself off at random times (like Wednesday afternoon), and doesn't come back up. You can't record anything if it's not powered on.
- Let's say you want to watch something that it's currently recording. Like, say, an episode of Farscape that's been running for half an hour, but hasn't finished yet. When you tell the box to start watching the show, it dumps you at the END of the recorded portion (aka LIVE TV).
- That, we can fix. Just rewind back to the beginning of the show. Slight hassle, but not horrible. Then, when the RECORDING ends (irrespective of where in the program YOU are), it dumps you back to the live feed. Not horrible for regular programs, but it sure sucks when you accidently see the final score of the basketball game you were watching.
- So you're halfway through a show, and you go run an errand. While you're gone, your SO watches something else. When you return to watch the rest of your show, the TW box starts at the BEGINNING, not where you stopped watching.
- TiVo has a function where you can record beyond the end of your show. College hoop, for instance, tends to run long, so you can tell it to record an extra half hour to make sure you get the end (and OT, if applicable). The TW has a similar function that you can program, but it doesn't work.
TiVo does all of the above admirably, with a user interface my technology stunted inlaws can use. That $500 never made so much sense, and the TW box is going back when I get a spare moment to do it.I don't do this for a living, but it seems that $/MIPS is the only benchmark even worth discussing, so shouldn't one be able to put together massive clusters of boxen to do the same thing, only without the SGI price tag?
Correct me, because I'm almost certian I'm wrong.
I hate to admit this, but I did an NPV analysis of the various payment schedules. This was back when monthly was $10, lifetime was $200, and (now unavailable) yearly was $100.
The upshot was that monthly was cheaper until month 17-18, and lifetime was cheaper after that. Yearly never made sense. So, if you expect to keep your TiVo longer than 18 months, which plan to get is pretty clear.
It's been pointed out that I should've included an end value for the lifetime subscription. I didn't, but that would just make the expected usefulness of the monthly plan even shorter.
I haven't re-run the numbers for the new pricing scheme, but I expect they'd be similar.
Not really. Your best bet might be keeping track of the Freevo project, although I've no experience with that, and it doesn't appear to be able to tape at the moment.
...some company sifting through my TV viewing habits
> TV listings are hardly worth $5/month...
We own two TiVos, and you're not just paying for the listings, but for the superb user interface. As many other posters have already mentioned:
"Yes, you COULD do everything TiVo does with a homegrown computer, but TiVo comes in a squeaky clean WORKING package, as opposed to cobbling something together from assorted parts lying around your garage."
I've had two SW upgrades during the couple of years we've had the TiVos, and both upgrades have brought features that significantly enhanced the use of my TiVo. Easily worth $10/month, IMHO.
>
TiVo are very open about what they track and what they don't. IIRC, they don't even trace back to the user, but use your information in conjunction with others to create profiles.
At that level, I don't have much of a problem with it.
During the Final Four last year, my beloved Kansas Jayhawks were playing. I came home from a happy hour the Friday night before the final four to find that TiVo had recorded the Final Four practices for all four teams. I didn't even know anybody would be nuts enough to cover that non-event. Needless to say, I was thrilled to see KU and their opponent's practices appear, unbidden, on my TiVo.
For that occasional miracle, I'll take all the Univision soap operas, shopping channel dreck, and Korean news I can delete, and I'll thank TiVo every time for trying.
I seem to recall people having trouble with the speed of the "now showing" window after their upgrade to 140hrs. If that's true, I can't imagine what 700 would do...
Anybody experienced this firsthand?
I once saw a spec (back in my VB programming days) that had something like four or five parameters in the hungarian notation before you even got to the variable name. something like
intLocalFnnameModulenameX
THAT gets a little absurd, IMHO. Not arguing that intX, or even intLocalX isn't useful, but you can twist yourself around an axle pretty damn quickly with this stuff.
have them "use it like it'll get used" for a while. Note all the problems, fix them. From there, you can create use cases that'll let sobody else do the testing.
It's CRITICAL, IMHO, that the people requesting the application get directly involved with how the front ends should work. If they don't, you're just asking for UI rework pain.
No market research, just raw ownership of the airwaves and, by extension, the markets.
R.I.P. Rev 105 in Mpls. We knew not what we had.
And don't give me 'from a friend who heard from a friend' etc., most likely that chain, however long, started with an ad.
I dunno, did you ever see a TiVo commercial? I think there were something like three, and they all sucked. TiVo's marketing has been horrible to date, IMHO.
Touche. OTOH, that perspective makes life into a big, dull distributed.net problem, and ignores what Kuhn (IIRC) called revolutionary science. So you set the rules to reflect our notion of reality now. You can get the bot to work through all sorts of permutations of the variables you assign it, but it won't come up with anything that's actually original.
If your reality didn't reflect that the earth goes around the sun, for instance, the bot couldn't do anything to disprove this. It would simply take the world it was presented, and maximize its reaction. I guess I'm wondering about what happens regarding breakthroughs that don't follow rulesets.
any opinions regarding intro to AI books?
I'm no rocket scientist, but I can't get past the notion that AI simply takes a goal, gives the algorithmic rules that apply to the world, and lets the algorithm go nuts trying out new stuff.
I'm thinking particularly of a genetic model I saw a few years ago, where the goal was "maximize speed," the ruleset provided physical characteristics of the world (i.e. gravity, friction coefficient of the ground, and so on), and while the results were interesting, I'd have trouble characterizing any of that as thought.
As such, when you set a goal of "reasonable conversation," and provide a ruleset and knowledge base, the machine isn't so much "thinking" as internally contesting two reactions to the ruleset.
Am I missing something?
Microsoft Mktng Geek: "Hi, this is Microsoft. We'd like to buy a booth at LinuxWorld."
LinuxWorld Sales Geek: "Damn kids, quit calling here. I don't know who the hell you're trying to amuse. It's not funny, and you're wasting my time."
You KNOW this happened.
Bantha poodoo. The video's going to have to go through a TV, to keep the price down in the markets you discuss. If you'd prefer to say "the number of internet capable machines will double by 2008," you've got a shot, as the internet ready PS2/Xbox/GCube/CableBox is where the battle has been for the last three years anyway, but they sure as heck won't be PCs.
over the period 1975-1978. If this trend continues....
"Ayyyyyy"
I seriously doubt we'll be seeing computer sales the way we used to in the future.
apologies to the simpsons.