That strikes more at the heart of the issue here than you may realize. The actual aircraft sitting in their hands is much closer to a compiled binary than source. You can poke at it, run it, look inside and try to reverse engineer it, but the real secret sauce that goes into making drones like this is the design/manufacturing techniques and massive high tech industrial base that are necessary to produce the components. The aircraft's engine isn't likely going to give up the secrets of directional crystal growth that go into manufacturing the turbine blades, and the camera's CCD isn't likely to yield the secrets of semiconductor fabrication necessary to produce another one.
I was in your situation about 7 years ago myself, was visiting RPI, applied to MIT as my long shot. Got into MIT somehow, got my SB MechE, MS MechE (had the chance to go PhD, decided not to), working in industry now and I think I've got a little bit of perspective on the engineering school experience.
From working directly with lots of engineers, helping profs select grad students, my own job search, and helping hire engineers there's a couple things you can do that'll greatly help your success no matter where you go to undergrad. Be passionate about something, stay focused, and do substantial work on it. Whether it's Robocup, Formula SAE, rocketry club, etc, just do it and make an impact. Do design, learn, work with people, and have fun doing it. Add being friendly and reliable (don't underestimate these!) to that and you really can go anywhere.
I think I could have gotten a similar education at most reputable engineering schools, everyone has the same statics and dynamics equations, but if you're passionate about what you do a top school can be a truly exhilarating experience. Besides getting the opportunity to do substantial work during undergrad for top PIs, there was just an intense forward current that came from living, working, and keeping up with brilliant people. Make sure you've got the humility to deal with people smarter than you and can take failures in stride and you really can come out with a tremendous life experience.
Whether you do Aero+MechE, or Aero or MechE will really depend on the specific program where you eventually go. When you end up working on real projects you'll likely find the impact is small, and from what I've seen Aero and MechE resumes are often thrown into the same pile. Both programs should be teaching structures, dynamics, design, controls, thermo/fluids (different spin, but it's the same math) and you may find your time is better spent on your own independent study than the specific requirements of a double degree.
You might not have noticed, but we've got this thing called the US Constitution. It says that the President is the Commander In Chief of the US Armed Forces. He answers to nobody on Manning's treatment. Say what you will about Manning's choice to cede his rights upon joining the armed forces, it's still terribly unbecoming for a nation that professes due process to allow this situation to happen. If it's all sensationalist lies about his treatment then it wouldn't do any harm to show that they're false either.
You're suggesting that the President doesn't have adequate control over the military. If that's the case, I'd say he's doing a pretty bad job. If he can't stand up for a Private how can he wrangle bigger issues?
Have you even looked at Nokia's stock price over the last 10 years? The stock dropped 20% on the Microsoft announcement after dropping 70% from its high in 2007. And that's after being down 32% from its highs in 2000. The company was sinking well before Microsoft came into the picture.
You've never wished you could have someone you're helping just show you their computer screen instead of trying to incorrectly describe it (or god forbid trying to walk them through setting up a remote connection)? A shot of the broken faucet they're bothering you about? It would cut a large part of the guesswork out of my random tech calls. Anyone who's a go-to technical person should see the value in the video chat cell phone form factor. Picture text messages work too, but the interactive nature of video chat really speeds things up.
If anyone here is actually interested in the science behind this you should have a look at some of the lab's publications on the subject. As per Slashdot tradition, all the brilliant points brought up so far in the comments already have answers, they're just a little bit harder to find this time.
Why does everyone keep saying investors will be angry as a fact when as this has been happening Google's stock price hasn't budged. In fact, the only thing that's made it move lately is the signal that the founders (the guys you seem to think will be bludgeoned) are going to be selling stock to bring their stake to less than 50%.
Considering how little we know about the emergence of intelligence from networks how is it possible to claim outright that an ANN can't approach the capabilities of a human brain? Real neurons are vastly more complex and aren't accurately modeled with such simple systems, but we don't have any idea what those complexities have to do with intelligence, so it seems to be quite the leap of faith to make claims on the topic.
Care to suggest how they differentiate between the thousands of applicants with both grades and standardized testing scores smashed up against the limits of the scales? Along that point, how do you pick the kid who's going to make MIT look good rather than hiding out in a room in Baker for four years? They need to lean heavily on the more subjective portions of the application like the essays and work portfolios in order to get any sort of meaningful picture of the applicant. That's also why this move makes perfect sense, splitting up the essay gets them a view from different angles without sacrificing any depth. After all, the 500 word essays didn't have any depth to begin with, and a 125 word essay is less likely to get polished to death by outside help.
I'm always confused as to why people get hung on this point so often. Why would someone in 1000 years (barring some apocalyptic situation), or even 20 years need a specific player to read a DVD, floppy disk, hard disk, or anything? All of these can be examined with more generic laboratory inspection equipment now, why is it unrealistic that 10 years from now you might have an optical disk scanner that reads just about anything? Even the encoding that the disks use isn't very complicated, we crack much more difficult codes all the time.
Re:Remember when Apple was going to buy Nintendo?
on
Apple Eyeing EA?
·
· Score: 1
You might want to look up the meaning of 'small fry' too. The term literally was correctly applied.
Re:Remember when Apple was going to buy Nintendo?
on
Apple Eyeing EA?
·
· Score: 4, Informative
How big the video game industry is? If you take 20 seconds to look it up you'd see that EA has a market cap of 6.49 billion vs Apple's market cap of 115.25 billion. Where do people keep getting the idea that EA is so big, they're literally a small fry compared to Apple.
When getting started developing something usually the first thing people do is run a patent search to see if the idea is unencumbered enough even bother with. If you've got a patent sitting there most engineers are going to stop in their tracks, whatever your actual feelings about licensing it are. In my experience with engineers who are actually doing the development your patent will scare them away and chances are that your idea will languish until the patent period expires. Maybe if you can somehow fit into the claims the fact that your intend it to be public domain that would work.
I'm a Real Engineer with a good bit of auto industry experience (though not a Chartered Engineer or PE as we tend to call them in the US, that's more for the civil engineering types), and I'm not sure where you're getting the idea that engineers aren't just 'jumped up mechanics'. Most of the best engineers I've worked with are captivated with experimentation, elegant design, and high performance applications. The best mechanics work with the same drive. The best engineers I know ARE jumped up mechanics. Just because someone learned how to analyze stresses and work with Navier-Stokes doesn't make them some zombie with a calculator.
As an engineer I'm not amazed by the fact that someone spent a year trying to find the cheapest plastic to mold a barely adequate oil pan from. We have a pretty good idea of what is cheap and what works. Tata worked with a different set of specifications than automotive designers in other countries and something different came out. I'm not ready to make a judgment whether the specs are wise or not, we'll see that over the next few years.
First Solar has a 25 year warranty on the power output of their panels. As far as I've seen this is pretty common for panels intended for large installations.
I think my point applies to what you're talking about very well. If qwerty had been patented we might be using a better keyboard layout today rather than dealing with the momentum that qwerty has developed. I agree that the pickup will definitely be slower, but after the initial ramp up we'd have a better system for ourselves because of the forced diversity. Would the loss because of waiting a few years for multitouch schemes to fight it in smaller user groups outweigh the benefits of the eventual standardization on a better evolved scheme?
Also, I'm not convinced 'natural' gestures are the best way to do multitouch. When training for various sports it seems that the best techniques are those that don't seem natural at all but have been proven more effective. Things like pitching a baseball or swinging a golf club come to mind immediately.
On the other hand, the problems that patents present to progress along a line of design can actually work in our favor. I've run up against patents in the past and in working on an alternate way of solving the problem I run into a better solution. There's no reason that what was patented is the best solution, it's most likely just the most obvious. It's actually a cool little trick for forcing development out of local minima (assuming a cost function on optimality like all sane people do).
Random rant: I've found 90% of the patents I run into are stuff someone patented to sit on and aren't actively developing. Apparently actually making the thing and marketing it are too hard, it's much more efficient to patent a swath of bad ideas and try to force licenses upon those who actually want to make progress.
The funny thing about this is that however many geeks there are that think it'd be fun to set up a jammer there's as many geeks out there who'd like nothing more than to track them down. I can see amateur radio operators having a field day (pun intended) hunting them down and helping the FCC hand out fines. No doubt crushing fines both because of the implications for emergency handling and because it's a strike against the telecoms. Tracking down cell phone jammers could become a major sport for radio operators if they become more common.
You sure of that? All the new electrical engineers I know (quite a few) hold a ton of respect for all the crazy things that were done with vacuum tubes back in the old days. The mechanical engineers (of which I am one) all love the old mechanical computers. Just because the new stuff is newer doesn't make the old stuff quaint or simple, in fact, most of the advances we have have made our jobs easier (and all of us less badass engineers). Those guys were taking on unbelievable problems with green CRTs, slide rules, and drafting tables.
The fact that those things exist, let alone work, should leave you slack-jawed.
It only looks like hammered dog meat if you don't know what you're looking at. I'm sure all the engineers that see the stuff are both amazed by the audacity of most of these designs and by the fact that they ever even approached the reliability they have with such complexity. On the other hand, I'm sure most of the same engineers have gripes about almost all of the design details.
You've still got to admire the complete absurdity of such machines though.
Hey there, I'm the guy that ended up designing and building the final electronics system for the Purple team. Did you get a chance to see the system I ended up doing? I was able to get some pretty good performance, reliably detecting a single infrared LED out to about 40 feet in an extremely noisy environment.
Slashdot Mirror
That strikes more at the heart of the issue here than you may realize. The actual aircraft sitting in their hands is much closer to a compiled binary than source.
You can poke at it, run it, look inside and try to reverse engineer it, but the real secret sauce that goes into making drones like this is the design/manufacturing techniques and massive high tech industrial base that are necessary to produce the components. The aircraft's engine isn't likely going to give up the secrets of directional crystal growth that go into manufacturing the turbine blades, and the camera's CCD isn't likely to yield the secrets of semiconductor fabrication necessary to produce another one.
I was in your situation about 7 years ago myself, was visiting RPI, applied to MIT as my long shot. Got into MIT somehow, got my SB MechE, MS MechE (had the chance to go PhD, decided not to), working in industry now and I think I've got a little bit of perspective on the engineering school experience.
From working directly with lots of engineers, helping profs select grad students, my own job search, and helping hire engineers there's a couple things you can do that'll greatly help your success no matter where you go to undergrad. Be passionate about something, stay focused, and do substantial work on it. Whether it's Robocup, Formula SAE, rocketry club, etc, just do it and make an impact. Do design, learn, work with people, and have fun doing it. Add being friendly and reliable (don't underestimate these!) to that and you really can go anywhere.
I think I could have gotten a similar education at most reputable engineering schools, everyone has the same statics and dynamics equations, but if you're passionate about what you do a top school can be a truly exhilarating experience. Besides getting the opportunity to do substantial work during undergrad for top PIs, there was just an intense forward current that came from living, working, and keeping up with brilliant people. Make sure you've got the humility to deal with people smarter than you and can take failures in stride and you really can come out with a tremendous life experience.
Whether you do Aero+MechE, or Aero or MechE will really depend on the specific program where you eventually go. When you end up working on real projects you'll likely find the impact is small, and from what I've seen Aero and MechE resumes are often thrown into the same pile. Both programs should be teaching structures, dynamics, design, controls, thermo/fluids (different spin, but it's the same math) and you may find your time is better spent on your own independent study than the specific requirements of a double degree.
YMMV and such. Good luck!
You might not have noticed, but we've got this thing called the US Constitution. It says that the President is the Commander In Chief of the US Armed Forces. He answers to nobody on Manning's treatment. Say what you will about Manning's choice to cede his rights upon joining the armed forces, it's still terribly unbecoming for a nation that professes due process to allow this situation to happen. If it's all sensationalist lies about his treatment then it wouldn't do any harm to show that they're false either.
You're suggesting that the President doesn't have adequate control over the military. If that's the case, I'd say he's doing a pretty bad job. If he can't stand up for a Private how can he wrangle bigger issues?
Have you even looked at Nokia's stock price over the last 10 years? The stock dropped 20% on the Microsoft announcement after dropping 70% from its high in 2007. And that's after being down 32% from its highs in 2000. The company was sinking well before Microsoft came into the picture.
You've never wished you could have someone you're helping just show you their computer screen instead of trying to incorrectly describe it (or god forbid trying to walk them through setting up a remote connection)? A shot of the broken faucet they're bothering you about? It would cut a large part of the guesswork out of my random tech calls. Anyone who's a go-to technical person should see the value in the video chat cell phone form factor. Picture text messages work too, but the interactive nature of video chat really speeds things up.
Cambridge is too expensive, we all moved out to Somerville.
If anyone here is actually interested in the science behind this you should have a look at some of the lab's publications on the subject. As per Slashdot tradition, all the brilliant points brought up so far in the comments already have answers, they're just a little bit harder to find this time.
Our research group's website:
http://groups.csail.mit.edu/locomotion/index.html
On the actual perching work:
http://groups.csail.mit.edu/robotics-center/public_papers/Cory08.pdf
http://groups.csail.mit.edu/robotics-center/public_papers/Hoburg09a.pdf
http://groups.csail.mit.edu/robotics-center/public_papers/Moore09.pdf
http://groups.csail.mit.edu/robotics-center/public_papers/Roberts09.pdf
Rick's PhD thesis on the subject:
http://groups.csail.mit.edu/locomotion/perching_media/CoryThesis.pdf
and on the controls side:
http://groups.csail.mit.edu/robotics-center/public_papers/Tedrake09a.pdf
Why does everyone keep saying investors will be angry as a fact when as this has been happening Google's stock price hasn't budged. In fact, the only thing that's made it move lately is the signal that the founders (the guys you seem to think will be bludgeoned) are going to be selling stock to bring their stake to less than 50%.
Considering how little we know about the emergence of intelligence from networks how is it possible to claim outright that an ANN can't approach the capabilities of a human brain? Real neurons are vastly more complex and aren't accurately modeled with such simple systems, but we don't have any idea what those complexities have to do with intelligence, so it seems to be quite the leap of faith to make claims on the topic.
Care to suggest how they differentiate between the thousands of applicants with both grades and standardized testing scores smashed up against the limits of the scales? Along that point, how do you pick the kid who's going to make MIT look good rather than hiding out in a room in Baker for four years? They need to lean heavily on the more subjective portions of the application like the essays and work portfolios in order to get any sort of meaningful picture of the applicant. That's also why this move makes perfect sense, splitting up the essay gets them a view from different angles without sacrificing any depth. After all, the 500 word essays didn't have any depth to begin with, and a 125 word essay is less likely to get polished to death by outside help.
SSH you!
I'm always confused as to why people get hung on this point so often. Why would someone in 1000 years (barring some apocalyptic situation), or even 20 years need a specific player to read a DVD, floppy disk, hard disk, or anything? All of these can be examined with more generic laboratory inspection equipment now, why is it unrealistic that 10 years from now you might have an optical disk scanner that reads just about anything? Even the encoding that the disks use isn't very complicated, we crack much more difficult codes all the time.
You might want to look up the meaning of 'small fry' too. The term literally was correctly applied.
How big the video game industry is? If you take 20 seconds to look it up you'd see that EA has a market cap of 6.49 billion vs Apple's market cap of 115.25 billion. Where do people keep getting the idea that EA is so big, they're literally a small fry compared to Apple.
When getting started developing something usually the first thing people do is run a patent search to see if the idea is unencumbered enough even bother with. If you've got a patent sitting there most engineers are going to stop in their tracks, whatever your actual feelings about licensing it are. In my experience with engineers who are actually doing the development your patent will scare them away and chances are that your idea will languish until the patent period expires. Maybe if you can somehow fit into the claims the fact that your intend it to be public domain that would work.
I'm a Real Engineer with a good bit of auto industry experience (though not a Chartered Engineer or PE as we tend to call them in the US, that's more for the civil engineering types), and I'm not sure where you're getting the idea that engineers aren't just 'jumped up mechanics'. Most of the best engineers I've worked with are captivated with experimentation, elegant design, and high performance applications. The best mechanics work with the same drive. The best engineers I know ARE jumped up mechanics. Just because someone learned how to analyze stresses and work with Navier-Stokes doesn't make them some zombie with a calculator.
As an engineer I'm not amazed by the fact that someone spent a year trying to find the cheapest plastic to mold a barely adequate oil pan from. We have a pretty good idea of what is cheap and what works. Tata worked with a different set of specifications than automotive designers in other countries and something different came out. I'm not ready to make a judgment whether the specs are wise or not, we'll see that over the next few years.
Valedictorian in EE at MIT? I wasn't aware we had valedictorians, maybe they got rid of it between then and now.
First Solar has a 25 year warranty on the power output of their panels. As far as I've seen this is pretty common for panels intended for large installations.
Here's a couple random links to back that up:
https://energy.wesrch.com/User_images/Pdf/L02_1221963706.pdf
http://www.evergreensolar.com/upload/pdf/us/Warranty_Cedar_Spruce_v1.5-060329_US.pdf
You sure? My LG Keybo (aka enV2) most definitely has a standard USB micro B. I'm looking at the cable right now.
I think my point applies to what you're talking about very well. If qwerty had been patented we might be using a better keyboard layout today rather than dealing with the momentum that qwerty has developed. I agree that the pickup will definitely be slower, but after the initial ramp up we'd have a better system for ourselves because of the forced diversity. Would the loss because of waiting a few years for multitouch schemes to fight it in smaller user groups outweigh the benefits of the eventual standardization on a better evolved scheme?
Also, I'm not convinced 'natural' gestures are the best way to do multitouch. When training for various sports it seems that the best techniques are those that don't seem natural at all but have been proven more effective. Things like pitching a baseball or swinging a golf club come to mind immediately.
On the other hand, the problems that patents present to progress along a line of design can actually work in our favor. I've run up against patents in the past and in working on an alternate way of solving the problem I run into a better solution. There's no reason that what was patented is the best solution, it's most likely just the most obvious. It's actually a cool little trick for forcing development out of local minima (assuming a cost function on optimality like all sane people do).
Random rant: I've found 90% of the patents I run into are stuff someone patented to sit on and aren't actively developing. Apparently actually making the thing and marketing it are too hard, it's much more efficient to patent a swath of bad ideas and try to force licenses upon those who actually want to make progress.
The funny thing about this is that however many geeks there are that think it'd be fun to set up a jammer there's as many geeks out there who'd like nothing more than to track them down. I can see amateur radio operators having a field day (pun intended) hunting them down and helping the FCC hand out fines. No doubt crushing fines both because of the implications for emergency handling and because it's a strike against the telecoms. Tracking down cell phone jammers could become a major sport for radio operators if they become more common.
You sure of that? All the new electrical engineers I know (quite a few) hold a ton of respect for all the crazy things that were done with vacuum tubes back in the old days. The mechanical engineers (of which I am one) all love the old mechanical computers. Just because the new stuff is newer doesn't make the old stuff quaint or simple, in fact, most of the advances we have have made our jobs easier (and all of us less badass engineers). Those guys were taking on unbelievable problems with green CRTs, slide rules, and drafting tables.
The fact that those things exist, let alone work, should leave you slack-jawed.
It only looks like hammered dog meat if you don't know what you're looking at. I'm sure all the engineers that see the stuff are both amazed by the audacity of most of these designs and by the fact that they ever even approached the reliability they have with such complexity. On the other hand, I'm sure most of the same engineers have gripes about almost all of the design details.
You've still got to admire the complete absurdity of such machines though.
Hey there, I'm the guy that ended up designing and building the final electronics system for the Purple team. Did you get a chance to see the system I ended up doing? I was able to get some pretty good performance, reliably detecting a single infrared LED out to about 40 feet in an extremely noisy environment.