You are absolutely right on the PDX-10, I stand corrected. Thank you: at least I embarrassed myself in front of Slashdot instead of a client. I had no idea there was a native 16:9 camera that cheap.
1/2" chips: I used to love the PD-150 with 1/2" chips, but I'd find it hard to go back now that I've tried 2/3" - much easier to muddy the background.
While I'm here: I've done 30p (and interpolated it from 60i via software, and 25p PAL, and so on), and I have to say that while 30p is most of the way there from 60i, 24p is still superior to my eye. I've used them side by side in a documentary - 24p with pulldown and 30p without - and 24p just feels better to me, even if it's staying on video. This was with both interviews and fast sports action.
For what it's worth, I'm told this is cultural, and that the Japanese always prefer 60fps, progressive or interlaced, for that sharp look. So YMMV.
Posting late (because I was out yesterday DPing a short film with a DVX100), but it's been ages since a subject has come up where I feel so well qualified to pipe up, so what the hell.
1) I agree with parent, the DVX100 (or 100A) is what you want. Don't argue with me, I've shot with most of the other cameras people are mentioning here, and it's one of those cases where I can recommend one model without qualification. Mostly because...
2)...24p trumps all other concerns at the prosumer level, assuming you want your results to "look like film". Go to 2-Pop or http://www.cinematography.net/, where such issues have been addressed dozens of times by people who know *exactly* what they're talking about. The DVX100(A) also has great color, a great lens, good sound, a large LCD, manual *and* servo focus, and a good line of accessories.
3) Get an anamorphic adaptor, if you have any intention of blowing up to film or projecting on a large screen. Research 4:3, 16:9 native, 16:9 squash, and anamorphic.
4) If resolution is really your bag, remember, prosumer HDV is around the corner (i.e., November for Sony's 3-chip, I believe). JVC's HDV is already available, but falls short for various reasons, mostly through being a 1-chipper. I wouldn't give up 24p to go HDV, though.
5) If your goal is to make a presentable short film, a good microphone is just as important as the camera. Bad sound kills even more amateur films than bad picture, and audiences are particularly unforgiving with sound. Get the nicest Sennheiser you can afford.
6) You also need to light, but there are lots of ghetto solutions to that, and to moving the camera. For Jebu's sake, get a good fluid tripod head. Do your research, find out why filmmakers usually have a wheelchair in their garage.
7) Re. PDX-10 and XL2: I'm virtually certain the PDX-10 has 16:9 squash, not native (16:9 squash is one of the few features the 100A has that the 100 doesn't, another being autofocus in 24p mode, btw.) The XL2 does look interesting, and the choice of lenses is a big draw as it was for the XL1, but it's getting some thumbs down over aliasing, color, and ergonomics. Personally I couldn't get enthused over the XL1, but I only shot with it for a day.
8) The underwater work probably shouldn't drive your decision, you can rent a rig for that. Or find a pool with a window, half-submerge a large fishtank, shoot dry for wet, etc. That's the fun part, being innovative enough to do things you shouldn't be able to do with the available equipment.
9) One last note: while I support the idea that you need to own a camera to experiment with, there's much to be said for renting a different camera for each project, too. No one camera has it all.
At the Borders near me they have a setup with a pair of headphones and an infrared bar code scanner. You can pick any CD off the shelf, scan it, and can then listen to a minute or so of any track. I believe the system imposes a very faint intermittent beep over the audio, too, for obvious reasons--I couldn't hear it on all tracks.
It made for a very pleasant and educational half hour of music browsing for me (annoying, though, that about 2/3 of the stations were mal/nonfunctioning).
Michael Kandel did that and all the best of Lem's complex meta-fiction translations (One Human Minute, Imaginary Magnitude, etc.). I agree, he's brilliant.
By way of introduction: I developed the core coarticulation and other algorithms for lip synching when I worked at a now-defunct company called...wait for it...LIPSinc. We thought the resulting lip synching was pretty damn convincing, so on my own I tested out our stuff with a hearing-impaired friend, with mixed results. Anyway, I don't know a little about this stuff, I know a *lot* about it.
What these guys have done is map phonemes onto exaggerated visemes (the pictures of the mouth). Not a bad idea at all! Bunch of problems, though. First, there's a data data reduction of about 3x in going from sound to video--there are 40-50 distinguishable phonemes, and 9-16 distinguishable visemes, depending on how you count each. This is because the visible part of the face only makes up the end of the vocal tract, a lot of distinctions between letters occurs without the involvement of the lips, like the difference between F and V, while others, like K, can be pronounced with the face in virtually any position. This is part of what makes lip reading so hard with a real person, and why they need a lot of context to pull it off. They also seem to be slowing down the timing, as if they recognized the phonemes and then synthesized each at the same length. This gives longer to recognize each one, but wrecks the visual prosody (rhthym) of the speech, which is a good cue for where the parts of speech are. Then there's the rest of the face. The eyebrows and head positions help you figure out key words, ends of clauses, tell if something is a question, etc.
Those who say that TTS is superior to lip reading have a point. Good TTS contains *more* accurate information than an uninterpreted stream of phonemes (itself 3x richer than a stream of visemes, as I said above), because the machine can do a Viterbi search to find the most likely sequence of words from a continuous stream of phonemes. Words also open up higher NLP functions, so you can do constraint relaxation to test whether "wreck a nice beach" or "recognize speech" fits better in the context.
Still, I'd like to see an experiment where the raw phonemes are fed, as text, to the recipient. I think with practice, your brain would start to decode the string (it manages with the sound, right?), despite the lack of word boundaries and the errors in phoneme detection (which is not all that high without text-I think seventy-something percent). Seems like an easier pattern recognition problem than lip reading. Who wants to go get funding?
This is great advice, in fact, I'll put trying out this pathway on my copious spare time list.
One meek suggestion: this kind of messing with the pipeline is fraught with peril, but scene conversion tools do exist, such as Okino's PolyTrans. The Maya obj exporter for Blender is production/stable, for that matter. So you could try modeling in the tool of your choice, then converting the scene to Maya format as a pre-process to the above.
If you're like me, you'll get tired of this in about two days, and go for an end-to-end package. If you already have Maya as part of your pipeline, well, it has a renderer in there already, and you'll need a good eye to find fault with it.
...Softimage XSI, Discreet (or whatever they call themselves this month) 3DSMax, Hash Animation Master. I always start out people new to 3D on Hash, it's cheap, full-featured, and has everything you need to learn the 3D animation trade (of which rendering is the least part), basically it kicks butt.
I've used 'em all, and personally I like the way Maya and SoftImage renderings look best, but far more important than the renderer is how good you are at lighting and texturing. If you take the time to build the material correctly and light it well, any of these tools will give you good results. Most of them are coming on, what, ten years old now?
IMO, the rendering engine comes into play when you really want photorealistic effects like global illumination or radiosity (and then only subtly), or special effects, like a plugin toon renderer...stuff you will no doubt have fun playing with, but aren't likely to impact your UI work much. Choose an interface and modeling tools you like, or spare your wallet, would be the criteria I'd recommend.
The authors decline to define postmodernism, for reasons of space. While I respect their decision, here's some insight from Frederick Jameson, William A. Lane Professor of Comparative literature and Director of the Graduate Program in Literature and the Center for Crirical Theory at Duke University, perilously near to where I live:
"Any sophisticated theory of the postmodern ought to bear something of the same relationship to Horkheimer and Adorno's old 'Culture Industry' concept as MTV or fractal ads bear to fifties television series."
If you don't know what this means, it's because your brain evolved to reject drivel. To be perfectly honest, I hope this is a hoax. Wouldn't be the first time.
But then, with postmodernism, you can't really tell the hoaxes from the honest nonsense.
Adam Gopnik of the New Yorker noted some time ago that the message of postmodern work is almost always trivial (like "violence is bad"), but couched in the most inscrutable and/or eye-catching terms (like "search for an interpretive skein within that overburdened word 'violence'" or "violence as style"). How about this one, from the paper: "Without a grand narrative, there will no be one common way to program, or even one common kind of interface between programs." More than one way to program? Sign me up for a grand narrative, post-haste!
I thought Slashdot was immune to this kind of idiocy. (Well...no, I didn't, but I can dream, can't I?)
Natural selection is easily falsifiable, as Darwin himself noted. One example: if an organism displays an adaptation which is of more benefit to some other organism than to itself--if horses evolved saddles--natural selection goes poof.
Natural selection doesn't go poof, though.
An aside: I consider it irresponsible that the story link is not balanced by a more mainstream view, and by mainstream I mean Google's opinion, not mine--why not the TalkOrigins site, which comes up first in Google on bombardier beetle evolution?
How much time did you spend talking to a live human on tech support at your favorite (or least favorite) company last year? These companies keep records--they know *exactly* how much you're worth to them. If a customer paid you $45/mo (most of which is not profit, remember), and cost you $33/min for tech support, what would you do? What would you do?
How Robert Cialdini, who has thought more about this kind of question than all of Slashdot put together, might break down the problem, according to the six principles set forth in his remarkable book "Influence, the Psychology of Persuasion":
Authority - This software comes from the acknowleged leader in the field. Reciprocity - We want people to pay us a ton of money for our own products. Consistency - We paid millions for our network, we're not willing to pay six figures for the software? Social Proof - Everybody else is buying this software, there must be a reason. Scarcity - It's the only solution in its price range, must be unique and valuable. Liking - They have the best (paid) salesmen.
Beat me to it. Coincidentally, I've been reading utopian novels for a month or so, including More's "Utopia" and Edward Bellamy's "Looking Backward". Both of these have, in my opinion, appalling authoritarian overtones, from the vantage of the early 21st century. Especially the latter, with talk of "mustering" people into the "industrial army". Yikes! But More's society required travel permits for leaving your neighborhood, and depended on slavery. Not so easy building a utopia.
But I digress. "The Dispossessed" gets mentioned in the same breath as these. I'd already read it three or four times growing up and since. I think the key feature which the above summary misses is that from which the title is taken: citizens of Annares do not acquire or keep personal possessions. The other world is more or less like ours, politically and economically. This was in LeGuin's heavy dualist period, shortly after "Left Hand of Darkness". It owes much to LeGuin's admiration for Paul Goodman.
For what its worth, every time I reread it, I find the language more beautiful and the human conflicts (whichever critic claimed it lacks them needs to read it again--or perhaps for the first time) more rivetting.
I think you're going to get a lot of people giving you the Kissinger line: "Academic politics are so vicious precisely because so little is at stake". I have a different take.
Last week I had lunch with a friend in the academic fold, to which I'm poised to return myself, and she complained with some rancour about the abundance of talentless hacks that cop credit and brown-nose their way to the top.
After four years with a VC startup (now being lowered into the earth) it all sounded quaint to me. I'd rather have talentless hacks stealing my work for a few years than watch the PHB lie his ass off to the board quarter after quarter without even a concept of shame, while the entire ill-conceived edifice crumbles around us all.
That is to say, go for it. Your reasons are exactly the ones I'd give, extrapolated a bit: I'd rather contribute in some infinitesimal way to the progress of science, however political or tedious the realities of research (who said "most of science is about as glamorous as ditch-digging", was it Asimov?), than help one more heinous moron pay off his SUV.
As for the money, I bet I'm not the only one here prepared for noble poverty, if such a thing still exists under the sun. Go, don't look back!
Well relax, your problems are solved. We'll have you talking like a native (again) before you can do ASL for "technology to the rescue!"
(With apologies to the entire population of Italy and those of Italian extraction, and, well, everyone everywhere. I'm just a dumb Mick, don't whack me!)
Never happen, but it's fun to think about. Throw a blur on those black lines on the map. "Well, you appear to live 70% in Israel and 30% in Palestine, please split your taxes, votes, political leanings, religious doctrines, prejudices, and so on accordingly."
My take: people live where they think they live. For tens of thousands of years, people have defined places using prepositional phrases. Now we can use coordinates, great. But if the numbers conflict with those definitions, it's the numbers that need adjusting.
Oddly enough, this is germain to my germaine to my half-baked, nowhere-near-ready-for-public-consumption personal project, which involves trying to represent places both with GPS coordinates and phrases like "down by the riverside".
(This has to be tongue-in-cheek, but since it's currently modded up as "insightful"...)
So, I get sued by my ex-employer if I use my experience, or fired by my current employer if I don't? The same current employer that forced me to sign an NDA/non-compete when they hired me (or, better yet, shortly thereafter)? Here in NC, I'm told that an employee manual you have never read can be interpreted as a legally binding document.
...in Player Piano, in 1955 or so. A recording, no matter how faithful, is a recording, it captures a performance, not a performer. The Disklavier is a player piano, a really good one. It makes the hammers hit the strings in (insert meaningless technical quibble here) the same way the pianist's fingers did. But until it captures Glenn Gould's humming, Stevie Wonder's head-bobbing, or Tori Amos' (essentially) fucking the piano stool, I will not be fooled into thinking that I am experiencing a live performance.
(Been playing since I was four, and I prefer Steinways to Yamahas, but that's another matter.)
The "100% vowel detection" claim sets off alarm bells for me. Sure, pure vowels tend to show up on the face, but there are lots of characteristics of speech which occur down in the throat, or back of the tongue...how do they plan to distinguish between sh, ch, and j? S and Z? F and V? For now, they don't.
I also just don't see how the claim can be accurate. I can say the "ih" phoneme with my jaw in any position, and I can say "a e i o" without moving my cheeks or jaw at all. What gives?
Human lip readers need *context*, and lots of it. This one I'll believe when I can use the demo myself.
As for losing subtleties of communication, I think the real problem is in synthesis. I work on the opposite side of this problem, generating lip movements from audio (i.e. lip synching). A lot of the subtleties you might think you'd lose are actually there in both signals, the audio and the muscles. For example, you can tell when somebody's smiling over the phone, the change in the shape of the mouth makes the phonemes sound different. Shouting invokes different muscles from normal speech, emphasis might be picked up from the eyebrows, and so on. But even if you detect such things on the face, no voice synthesis engine is capable of rendering the accompanying vocal effects.
I'm willing to believe that this is a new and sophisticated tool, though I see there are already a bunch of credible posts about prior art (I go to SIGGRAPH, too).
My issue is that a lot more than physics needs to go into an animation. There's brain up there controlling all those muscles, it's not just a bunch of sticks and rubber bands. A character animator is an *actor*, part of his or her job is to give the appearance of intention to a character's performance.
Falling down the stairs is (relatively) easy. Show me the panic the moment he realizes he's lost his balance and can't stop himself. Does he flop like a drunk or roll out like Jet Li?
You still need to control a character with a human brain, whether that brain belongs to a mocapped stunt man or an animator.
The technology of this article shifts a call from the top to the bottom of this list. They admit that the advance is not in AI or voice tech, but in making the experience "resemble a conversation". So at its best, this will still let grandma have *some* access to the information she could have had before from a live human. At its worst, it's a puppet show to distract us from the fact that we're not getting very good service.
Caveats:
a) I know nothing about genetics or law myself. I learned all this from the genetics law expert I sat next to on a plane last week.
b)The duration of the explanation was part of a flight from Salt Lake to Seattle
c) I had a first class upgrade and took full advantage of the free Heinekens. That is to say, I hope I'm remembering this right.
Goes like this. It's illegal to patent an object, right? But a sequence of DNA in addition to containing the gene you're interested in, is always full of random and irrelevant pairs. So what they want to patent is not the gene as it naturally occurs, with all the junk DNA in it, but a cleaned-up version containing only those bits which are relevant to the patent. This is not a naturally occuring sequence, and so is patentable. So to answer the fellow who says "wait, I have that gene, every cell of me is prior art," no, you don't have that exact gene, yours contains different randomness. Yes, this sounds like a legalistic dodge to me too, and the expert acknowleged the point, but there it is.
A further wrinkle is that they patent the transcriptase necessary to make the cleaned-up gene, not the gene itself, though I had a sufficient buzz by that point in the conversation that I was ready to talk about football.:)
"Our pile of patent trading cards is bigger than your pile of patent trading cards, (plus this one is, like, a Wizard, and really powerful!). Therefore this technology partnership goes *our* way, not yours..."
2) Company valuation.
This kind of thing still actually impresses investors: "Stupidco. is a recognized leader in stupidity, and holds key patents in the field."
3) Patent lawyers.
"You want to patent the moon? No, no, I don't think that's a bad idea at all! You did bring your check book, right? We may need to patent full, crescent and new moon separately..."
4) If we don't do it...
"You really want to take the risk that Idiots.com will patent idiocy before we can patent stupidity? Sure it's a dumb patent, but that doesn't mean they won't get it!"
5) Difficulty of challenging a patent in court.
Witness: "...important innovation relating to the doping of spherical semiconductors during non-contact processing in the liquid state..."
Slashdot Mirror
Doh! And of course by 1/2" I mean 1/4", and the DVX100's are 1/3". I will shut up now before I dig myself any deeper.
You are absolutely right on the PDX-10, I stand corrected. Thank you: at least I embarrassed myself in front of Slashdot instead of a client. I had no idea there was a native 16:9 camera that cheap.
1/2" chips: I used to love the PD-150 with 1/2" chips, but I'd find it hard to go back now that I've tried 2/3" - much easier to muddy the background.
While I'm here: I've done 30p (and interpolated it from 60i via software, and 25p PAL, and so on), and I have to say that while 30p is most of the way there from 60i, 24p is still superior to my eye. I've used them side by side in a documentary - 24p with pulldown and 30p without - and 24p just feels better to me, even if it's staying on video. This was with both interviews and fast sports action.
For what it's worth, I'm told this is cultural, and that the Japanese always prefer 60fps, progressive or interlaced, for that sharp look. So YMMV.
Posting late (because I was out yesterday DPing a short film with a DVX100), but it's been ages since a subject has come up where I feel so well qualified to pipe up, so what the hell.
...24p trumps all other concerns at the prosumer level, assuming you want your results to "look like film". Go to 2-Pop or http://www.cinematography.net/, where such issues have been addressed dozens of times by people who know *exactly* what they're talking about. The DVX100(A) also has great color, a great lens, good sound, a large LCD, manual *and* servo focus, and a good line of accessories.
1) I agree with parent, the DVX100 (or 100A) is what you want. Don't argue with me, I've shot with most of the other cameras people are mentioning here, and it's one of those cases where I can recommend one model without qualification. Mostly because...
2)
3) Get an anamorphic adaptor, if you have any intention of blowing up to film or projecting on a large screen. Research 4:3, 16:9 native, 16:9 squash, and anamorphic.
4) If resolution is really your bag, remember, prosumer HDV is around the corner (i.e., November for Sony's 3-chip, I believe). JVC's HDV is already available, but falls short for various reasons, mostly through being a 1-chipper. I wouldn't give up 24p to go HDV, though.
5) If your goal is to make a presentable short film, a good microphone is just as important as the camera. Bad sound kills even more amateur films than bad picture, and audiences are particularly unforgiving with sound. Get the nicest Sennheiser you can afford.
6) You also need to light, but there are lots of ghetto solutions to that, and to moving the camera. For Jebu's sake, get a good fluid tripod head. Do your research, find out why filmmakers usually have a wheelchair in their garage.
7) Re. PDX-10 and XL2: I'm virtually certain the PDX-10 has 16:9 squash, not native (16:9 squash is one of the few features the 100A has that the 100 doesn't, another being autofocus in 24p mode, btw.) The XL2 does look interesting, and the choice of lenses is a big draw as it was for the XL1, but it's getting some thumbs down over aliasing, color, and ergonomics. Personally I couldn't get enthused over the XL1, but I only shot with it for a day.
8) The underwater work probably shouldn't drive your decision, you can rent a rig for that. Or find a pool with a window, half-submerge a large fishtank, shoot dry for wet, etc. That's the fun part, being innovative enough to do things you shouldn't be able to do with the available equipment.
9) One last note: while I support the idea that you need to own a camera to experiment with, there's much to be said for renting a different camera for each project, too. No one camera has it all.
At the Borders near me they have a setup with a pair of headphones and an infrared bar code scanner. You can pick any CD off the shelf, scan it, and can then listen to a minute or so of any track. I believe the system imposes a very faint intermittent beep over the audio, too, for obvious reasons--I couldn't hear it on all tracks.
It made for a very pleasant and educational half hour of music browsing for me (annoying, though, that about 2/3 of the stations were mal/nonfunctioning).
Michael Kandel did that and all the best of Lem's complex meta-fiction translations (One Human Minute, Imaginary Magnitude, etc.). I agree, he's brilliant.
Posting late, but wtf.
By way of introduction: I developed the core coarticulation and other algorithms for lip synching when I worked at a now-defunct company called...wait for it...LIPSinc. We thought the resulting lip synching was pretty damn convincing, so on my own I tested out our stuff with a hearing-impaired friend, with mixed results. Anyway, I don't know a little about this stuff, I know a *lot* about it.
What these guys have done is map phonemes onto exaggerated visemes (the pictures of the mouth). Not a bad idea at all! Bunch of problems, though. First, there's a data data reduction of about 3x in going from sound to video--there are 40-50 distinguishable phonemes, and 9-16 distinguishable visemes, depending on how you count each. This is because the visible part of the face only makes up the end of the vocal tract, a lot of distinctions between letters occurs without the involvement of the lips, like the difference between F and V, while others, like K, can be pronounced with the face in virtually any position. This is part of what makes lip reading so hard with a real person, and why they need a lot of context to pull it off. They also seem to be slowing down the timing, as if they recognized the phonemes and then synthesized each at the same length. This gives longer to recognize each one, but wrecks the visual prosody (rhthym) of the speech, which is a good cue for where the parts of speech are. Then there's the rest of the face. The eyebrows and head positions help you figure out key words, ends of clauses, tell if something is a question, etc.
Those who say that TTS is superior to lip reading have a point. Good TTS contains *more* accurate information than an uninterpreted stream of phonemes (itself 3x richer than a stream of visemes, as I said above), because the machine can do a Viterbi search to find the most likely sequence of words from a continuous stream of phonemes. Words also open up higher NLP functions, so you can do constraint relaxation to test whether "wreck a nice beach" or "recognize speech" fits better in the context.
Still, I'd like to see an experiment where the raw phonemes are fed, as text, to the recipient. I think with practice, your brain would start to decode the string (it manages with the sound, right?), despite the lack of word boundaries and the errors in phoneme detection (which is not all that high without text-I think seventy-something percent). Seems like an easier pattern recognition problem than lip reading. Who wants to go get funding?
This is great advice, in fact, I'll put trying out this pathway on my copious spare time list.
One meek suggestion: this kind of messing with the pipeline is fraught with peril, but scene conversion tools do exist, such as Okino's PolyTrans. The Maya obj exporter for Blender is production/stable, for that matter. So you could try modeling in the tool of your choice, then converting the scene to Maya format as a pre-process to the above.
If you're like me, you'll get tired of this in about two days, and go for an end-to-end package. If you already have Maya as part of your pipeline, well, it has a renderer in there already, and you'll need a good eye to find fault with it.
...Softimage XSI, Discreet (or whatever they call themselves this month) 3DSMax, Hash Animation Master. I always start out people new to 3D on Hash, it's cheap, full-featured, and has everything you need to learn the 3D animation trade (of which rendering is the least part), basically it kicks butt.
I've used 'em all, and personally I like the way Maya and SoftImage renderings look best, but far more important than the renderer is how good you are at lighting and texturing. If you take the time to build the material correctly and light it well, any of these tools will give you good results. Most of them are coming on, what, ten years old now?
IMO, the rendering engine comes into play when you really want photorealistic effects like global illumination or radiosity (and then only subtly), or special effects, like a plugin toon renderer...stuff you will no doubt have fun playing with, but aren't likely to impact your UI work much. Choose an interface and modeling tools you like, or spare your wallet, would be the criteria I'd recommend.
"Any sophisticated theory of the postmodern ought to bear something of the same relationship to Horkheimer and Adorno's old 'Culture Industry' concept as MTV or fractal ads bear to fifties television series."
If you don't know what this means, it's because your brain evolved to reject drivel. To be perfectly honest, I hope this is a hoax. Wouldn't be the first time.
But then, with postmodernism, you can't really tell the hoaxes from the honest nonsense.
Adam Gopnik of the New Yorker noted some time ago that the message of postmodern work is almost always trivial (like "violence is bad"), but couched in the most inscrutable and/or eye-catching terms (like "search for an interpretive skein within that overburdened word 'violence'" or "violence as style"). How about this one, from the paper: "Without a grand narrative, there will no be one common way to program, or even one common kind of interface between programs." More than one way to program? Sign me up for a grand narrative, post-haste!
I thought Slashdot was immune to this kind of idiocy. (Well...no, I didn't, but I can dream, can't I?)
Natural selection is easily falsifiable, as Darwin himself noted. One example: if an organism displays an adaptation which is of more benefit to some other organism than to itself--if horses evolved saddles--natural selection goes poof.
Natural selection doesn't go poof, though.
An aside: I consider it irresponsible that the story link is not balanced by a more mainstream view, and by mainstream I mean Google's opinion, not mine--why not the TalkOrigins site, which comes up first in Google on bombardier beetle evolution?
How much time did you spend talking to a live human on tech support at your favorite (or least favorite) company last year? These companies keep records--they know *exactly* how much you're worth to them. If a customer paid you $45/mo (most of which is not profit, remember), and cost you $33/min for tech support, what would you do? What would you do?
How Robert Cialdini, who has thought more about this kind of question than all of Slashdot put together, might break down the problem, according to the six principles set forth in his remarkable book "Influence, the Psychology of Persuasion":
Authority - This software comes from the acknowleged leader in the field.
Reciprocity - We want people to pay us a ton of money for our own products.
Consistency - We paid millions for our network, we're not willing to pay six figures for the software?
Social Proof - Everybody else is buying this software, there must be a reason.
Scarcity - It's the only solution in its price range, must be unique and valuable.
Liking - They have the best (paid) salesmen.
Beat me to it. Coincidentally, I've been reading utopian novels for a month or so, including More's "Utopia" and Edward Bellamy's "Looking Backward". Both of these have, in my opinion, appalling authoritarian overtones, from the vantage of the early 21st century. Especially the latter, with talk of "mustering" people into the "industrial army". Yikes! But More's society required travel permits for leaving your neighborhood, and depended on slavery. Not so easy building a utopia.
But I digress. "The Dispossessed" gets mentioned in the same breath as these. I'd already read it three or four times growing up and since. I think the key feature which the above summary misses is that from which the title is taken: citizens of Annares do not acquire or keep personal possessions. The other world is more or less like ours, politically and economically. This was in LeGuin's heavy dualist period, shortly after "Left Hand of Darkness". It owes much to LeGuin's admiration for Paul Goodman.
For what its worth, every time I reread it, I find the language more beautiful and the human conflicts (whichever critic claimed it lacks them needs to read it again--or perhaps for the first time) more rivetting.
I think you're going to get a lot of people giving you the Kissinger line: "Academic politics are so vicious precisely because so little is at stake". I have a different take.
Last week I had lunch with a friend in the academic fold, to which I'm poised to return myself, and she complained with some rancour about the abundance of talentless hacks that cop credit and brown-nose their way to the top.
After four years with a VC startup (now being lowered into the earth) it all sounded quaint to me. I'd rather have talentless hacks stealing my work for a few years than watch the PHB lie his ass off to the board quarter after quarter without even a concept of shame, while the entire ill-conceived edifice crumbles around us all.
That is to say, go for it. Your reasons are exactly the ones I'd give, extrapolated a bit: I'd rather contribute in some infinitesimal way to the progress of science, however political or tedious the realities of research (who said "most of science is about as glamorous as ditch-digging", was it Asimov?), than help one more heinous moron pay off his SUV.
As for the money, I bet I'm not the only one here prepared for noble poverty, if such a thing still exists under the sun. Go, don't look back!
Well relax, your problems are solved. We'll have you talking like a native (again) before you can do ASL for "technology to the rescue!"
(With apologies to the entire population of Italy and those of Italian extraction, and, well, everyone everywhere. I'm just a dumb Mick, don't whack me!)
Never happen, but it's fun to think about. Throw a blur on those black lines on the map. "Well, you appear to live 70% in Israel and 30% in Palestine, please split your taxes, votes, political leanings, religious doctrines, prejudices, and so on accordingly."
My take: people live where they think they live. For tens of thousands of years, people have defined places using prepositional phrases. Now we can use coordinates, great. But if the numbers conflict with those definitions, it's the numbers that need adjusting.
Oddly enough, this is germain to my germaine to my half-baked, nowhere-near-ready-for-public-consumption personal project, which involves trying to represent places both with GPS coordinates and phrases like "down by the riverside".
(This has to be tongue-in-cheek, but since it's currently modded up as "insightful"...)
So, I get sued by my ex-employer if I use my experience, or fired by my current employer if I don't? The same current employer that forced me to sign an NDA/non-compete when they hired me (or, better yet, shortly thereafter)? Here in NC, I'm told that an employee manual you have never read can be interpreted as a legally binding document.
Am I a slave yet?
Bowed Piano Ensemble
...in Player Piano, in 1955 or so. A recording, no matter how faithful, is a recording, it captures a performance, not a performer. The Disklavier is a player piano, a really good one. It makes the hammers hit the strings in (insert meaningless technical quibble here) the same way the pianist's fingers did. But until it captures Glenn Gould's humming, Stevie Wonder's head-bobbing, or Tori Amos' (essentially) fucking the piano stool, I will not be fooled into thinking that I am experiencing a live performance.
(Been playing since I was four, and I prefer Steinways to Yamahas, but that's another matter.)
The "100% vowel detection" claim sets off alarm bells for me. Sure, pure vowels tend to show up on the face, but there are lots of characteristics of speech which occur down in the throat, or back of the tongue...how do they plan to distinguish between sh, ch, and j? S and Z? F and V? For now, they don't.
I also just don't see how the claim can be accurate. I can say the "ih" phoneme with my jaw in any position, and I can say "a e i o" without moving my cheeks or jaw at all. What gives?
Human lip readers need *context*, and lots of it. This one I'll believe when I can use the demo myself.
As for losing subtleties of communication, I think the real problem is in synthesis. I work on the opposite side of this problem, generating lip movements from audio (i.e. lip synching). A lot of the subtleties you might think you'd lose are actually there in both signals, the audio and the muscles. For example, you can tell when somebody's smiling over the phone, the change in the shape of the mouth makes the phonemes sound different. Shouting invokes different muscles from normal speech, emphasis might be picked up from the eyebrows, and so on. But even if you detect such things on the face, no voice synthesis engine is capable of rendering the accompanying vocal effects.
I'm willing to believe that this is a new and sophisticated tool, though I see there are already a bunch of credible posts about prior art (I go to SIGGRAPH, too).
My issue is that a lot more than physics needs to go into an animation. There's brain up there controlling all those muscles, it's not just a bunch of sticks and rubber bands. A character animator is an *actor*, part of his or her job is to give the appearance of intention to a character's performance.
Falling down the stairs is (relatively) easy. Show me the panic the moment he realizes he's lost his balance and can't stop himself. Does he flop like a drunk or roll out like Jet Li?
You still need to control a character with a human brain, whether that brain belongs to a mocapped stunt man or an animator.
2010: Homes made in prefabricated modules...guess he's never been to rural North Carolina.
2010: Orgasm by email. Oh, wait, we already have this. I'm reliably informed.
Also 2010: 25% of all TV personalities will be synthetic. Oh, wait...
Hey bein' one a them futurists is easy!
...keeping the customer from costing you any money.
CRM is *expensive*. Forrester Research did a study a while back on the average cost of handling customer calls by various means:
Telephone: $33.00/incident
Email: $9.99/incident
Chat: $7.80/incident
Message Boards: $4.57/incident
Knowledge Base: $1.17/incident
The technology of this article shifts a call from the top to the bottom of this list. They admit that the advance is not in AI or voice tech, but in making the experience "resemble a conversation". So at its best, this will still let grandma have *some* access to the information she could have had before from a live human. At its worst, it's a puppet show to distract us from the fact that we're not getting very good service.
Caveats:
:)
a) I know nothing about genetics or law myself. I learned all this from the genetics law expert I sat next to on a plane last week.
b)The duration of the explanation was part of a flight from Salt Lake to Seattle
c) I had a first class upgrade and took full advantage of the free Heinekens. That is to say, I hope I'm remembering this right.
Goes like this. It's illegal to patent an object, right? But a sequence of DNA in addition to containing the gene you're interested in, is always full of random and irrelevant pairs. So what they want to patent is not the gene as it naturally occurs, with all the junk DNA in it, but a cleaned-up version containing only those bits which are relevant to the patent. This is not a naturally occuring sequence, and so is patentable. So to answer the fellow who says "wait, I have that gene, every cell of me is prior art," no, you don't have that exact gene, yours contains different randomness. Yes, this sounds like a legalistic dodge to me too, and the expert acknowleged the point, but there it is.
A further wrinkle is that they patent the transcriptase necessary to make the cleaned-up gene, not the gene itself, though I had a sufficient buzz by that point in the conversation that I was ready to talk about football.
1) Patent swapping.
"Our pile of patent trading cards is bigger than your pile of patent trading cards, (plus this one is, like, a Wizard, and really powerful!). Therefore this technology partnership goes *our* way, not yours..."
2) Company valuation.
This kind of thing still actually impresses investors: "Stupidco. is a recognized leader in stupidity, and holds key patents in the field."
3) Patent lawyers.
"You want to patent the moon? No, no, I don't think that's a bad idea at all! You did bring your check book, right? We may need to patent full, crescent and new moon separately..."
4) If we don't do it...
"You really want to take the risk that Idiots.com will patent idiocy before we can patent stupidity? Sure it's a dumb patent, but that doesn't mean they won't get it!"
5) Difficulty of challenging a patent in court.
Witness: "...important innovation relating to the doping of spherical semiconductors during non-contact processing in the liquid state..."
Jurors: "?"