In common usage TSMC is the foundry not Apple (even if apple started buying equipment to process wafers, then it would simply be running a captive fab).
FWIW, the common terms of large foundry contracts have varied greatly over the last few decades. Some examples of "pure" pricing models:
* wafer starts (to reserve some fraction of capacity) * processed wafers (that had their wafer process monitor circuits working within a range of pre-agreed parameters). * working die (pre-diced chips that pass a short customer defined wafer probe tests)
Nearly all large contracts are a blend between these pure pricing models and are a function of the times.
When wafers are cheap and factories have lots of capacity (say when a new process comes on line, the older factories may have extra capacity and/or a competitive yield advantage relative to other companies), OR when the process is new and proper yield correlations are unknown, most customers can get the working die pricing model.
When capacity is really tight, a large customer might have to pay some amount of money for a wafer start just to reserve capacity at the factory (even if the processed wafer doesn't yield any working parts). Typically this somewhat negates most of the volume discount they would get and puts the pricing more on par with a smaller customer.
Of course everything is negotiable and there is probably a price for everything. However, given the current pricing environment with the capacity limitations TSMC had with 28nm, I doubt anyone could negotiate pure working die deal now unless they paid a handsome penny for each one.
Most folks in the (fab-less) semiconductor industry are referencing the foundry business model of TSMC. In the historical foundry business model, a foundry is where you can contract to have *custom* metal parts forged. In modern usage, SEH would generally be considered a supplier to TSMC, since the silicon wafer is more of a *standard* part, not a *custom* part.
SEH might have been considered analogous to foundry of sorts since a historical metal-works foundry poured metal into casts and finished them and SEH actually make the silicon ingots and cuts and finishes them into "blank" silicon wafers. However, since it is TSMC which sells custom services, and the wafers are more-or-less standard parts, TSMC are more akin to the foundry provider from a business sense and thus are called that in modern usage.
Fly by wires isn't the same as fly by wireless which what it would be if you didn't have pilots (unless they went the way of the old "TOW" missals which were connected by fiberoptic cable).
Given the possibilities of jamming and hacking, I think there is less of a chance of a flight deck issue than other ways of hacking/jamming...
The slippery slope is that perfecting this technique is a stepping stone to designer babies.
Apparently, the current technology allows replacement of the whole nucleus, allowing the nucleus of a fertilized egg w/ defective mitochondria to be placed into another fertilized egg with healthy mitochondria. The implication is that this fertilized egg is placed back in the mother (but it could be anyone). Many folks are pretty sure that we are pretty close to the ability to selectively replace a few chromosomes in that nucleus during the swap. Presumably, the first application of this would be to replace/remove a obviously defective chromosome (say trisomy-21 aka down's syndrome) which seems like it might be the first step down the slope.
As mappings of gene-expressions to chromosomes get more accurate, you might imagine that some specific chromosome could be selected for (take all the chromosomes except #k and get that from some other place). Although you might think that the next step might be making a kid with the "best" selection of chromosomes from a single mother and father, identifying a gene in a chromosome is currently a destructive procedure so that's not an easy path to take. Instead, the next logical step (if you allow for N>2 people in the procedure) is to pick a known good chromosome from a 3rd party. With enough 3rd parties involved, you pretty much have a designer baby...;^)
Of course there is an ethical side to this as well (gotta create/destroy a bunch of *human* fertilized eggs to do this) and that doesn't sit that well with many folks.
there are many, many desperate children that feel so hopeless and lonely right now in some orphanage
Adoption in many countries is very difficult, and plenty of potential parents do not qualify. My wife and I are financially secure, and are very successfully raising two of our own kids. But we had room in our home and our hearts for at least one more, and looked into adoption. We were turned down. The reasons given were that we were too old (I am in my 50s and my wife is in her 40s), and we already have kids of our own, and childless couples would be given priority.
If there really are orphanages full of desperate children, then governments are doing an incredibly poor job of matching them up with willing and capable parents.
Just think of using a donor egg as "adopting" that egg, and I don't think that's as difficult as adopting a baby in many countries (where IVF was available in the first place). The primary thing the 3-person IVF really enables over the commonly available 2-person IVF is that mother can have the option to be biologically related to the child if the mitochondrial dna in her eggs were somehow defective. Another way to look at it, you still have to "adopt" an egg w/ healthy mitochondria from someone (and if you were open to the normal adoption process in the first place, this would appear to be a don't care issue).
OAM multiplexing can not be implemented in the existing long-haul optical fiber systems, since these systems are based on single-mode fibers, which inherently do not support OAM states of light. Instead, few-mode or multi-mode fibers need to be used. Additional problem for OAM multiplexing implementation is caused by the mode coupling that is present in the fiber, making direct-detection OAM multiplexing still not being realized in long-haul communications. In some specialty fibers, OAM states were transmitted with 97% purity after 20 meters.
Basically this demonstration technique uses specially designed fibers that can carry the "donut" TEM mode required for OAM which is the reason they made a comment that the most likely for fibers the implement this technique "might be to install them to span the short distances between servers on giant 'server farms'"...
I think they just learned the lesson of Las Vegas. Vegas basically started out as an organized money skimming organization fronted by gambling operations (i.e., attempting to make more money by bypassing part the system). After a while, the proprieters found that once they figured out how to run the above board business, it was way more lucrative than the illegal part which means you might as well just concentrate your efforts on the above board business (or suffer the opportunity cost of not doing so).
Basically the lesson is that there's lots of money to be made out in the world by people with skills who get out of the small game and learn how to play the bigger game.
But it's not the WOPR! Now THAT would be the bit of movie gear to have!:-)
It would be, but if I recall correctly it was only a prop wired for lights. Still, it would fun to have that in a data center, especially if its one that people tour.
As I recall, that although the WOPR (like other movie computers of that era) was basically a prop with dancing stereo equalizer lights for cinematic effect, I think there was actually an Apple ][ inside controlling an experimental flat panel display on the side.
Although GEB can be an entertaining book, I don't think a course that had GEB as reading material was what the original writer had in mind for a prototypical humanities course***.
*** yes, I know there have been university courses built around GEB, but there have been university philosophy courses built around star trek too
The field of engineering is more like that of a technologist (learning and applying currently well accepted science to perform the job of creating things). A scientific field is more one that is building the base of scientific knowledge which often is incremental from currently well accepted science. Of course there is a big overlap between the two in practice, but when reflected back to undergraduate instruction level, they become suspiciously similar, so much so that it's really hard to distinguish. I would argue that at the undergraduate teaching level, there's not much of a difference between science and technology. All the topics have a "dogma" of some sort which has been decided to be taught by the masters of currucula often in a framework that reflects culture of field. Any so called labs or experiments have well understood answers (and students often are subjected to negative reinforcement if they stray off the beaten path) and are only designed to give a "flavor" of science (like a sad version of a cargo cult).
SO if all students are doing is surveying the history of field (which pretty much sums up undergraduate level instruction), it's pretty much all humanities courses people are taking at that level (since technology is part of the recorded culture of humanity). Probably the only real thing that distinguishs a "humanities" course from a "technology" or "science" course from a knowledge point of view is how recent the history is, what aptitudes/prerequisites are required to understand the material covered, and maybe perhaps how relavent the topic is to future employers...
FWIW, I basically came to this opinion many years ago when I was doing "alumni-calling" for free-pizza and beer. I rang up an alum who was a biologist that graduated 1 year before watson and crick got their nobel prize for discovering DNA. He was still practicing molecular biology, but pretty much had to catch up in grad-school on all stuff that was actually going on during undergrad (as opposed to what the undergrad classes covered). We chatted a while about what he thought about undergraduate level education in biology before this and although he thought he learned a small about the scientific method and the practice of real-world science, the bulk of his classes were pretty much pre-DNA dogma and none of them even mentioned topics around the 3-d structure of molecules or X-ray crystallography which would become his area of specialization (although he did have some exposure to that through talking to a few professors in the chemistry dept.). The undergraduate classes he remembered most, were in fact humanities classes since they had ideas that were more stable over time and he wished he would have taken more of them.
"Once the rockets are up, who cares where they come down, that's not my department," says Wernher von Braun
Or of course the antectodal fictional situations exist too..
Our studies indicate the weapon is totally useless in warfare. It's not intended for use in your kind of warfare. It's the perfect peacetime weapon. That's why it's secret. So it's both immoral *and* unethical? Yes.
.... Let the engineers figure out a use for it. That's not our concern. Maybe somebody already has a use for it, one for which it's perfectly designed.
Not that I totally disagree with the indoctrination aspect of many humanities programs, but citing an example about some people trotting out the allegory of Plato's cave isn't really much different than other people trotting out trite observations like Moore's law or Object Oriented Programming or Security through Obscurity because they heard about it somewhere and sadly misapplying them.
I think it's really human nature to be indoctrinated in some way or another as the amount of knowledged to be learned is tremendous and it's really hard to organize in a way that allows for communication or cooperation with others w/o resorting to referencing it to some sort of framework. Unfortunatly, since that framework for understanding is part of the "critical-infrastructure", there's a strong incentive for people with agendas to twist it to their desires. It is after all a human endeavor fraught with human weaknesses.
Centuries ago, the critical-infrastructure of common understanding was probably religion and that was argued by uninformed interpretations until wars broke out. Decades ago it was repeated with VI vs Emacs;^)
It's not, actually. There are already asymmetric crypto algorithms which are believed to be quantum-resistant. They are typically based on the hardness of vector problems in n-dimensional integer lattices, or problems that have been proven reducible to such problems such as learning with errors.
Interesting read, but it doesn't address the questions at hand: Are cryptography advances and computer advances hand-in-hand? I believe the answer to that question is still no. Are big crypto advances in a next generation public key algorithm limited by our knowledge? I believe the answer to that question is still yes.
For example, the McEliece-like crypto system (which seems suspiciously analogus to your LWE paper) did not appear to have any major advances for many years (probably because sending around huge matrices is a real bummer and attempts to reduces this reduced the security parameters). We also have no direct knowledge about its quantum resistance other than some assumptions. Who knows, but its security profile may be very entwined with algorithms for error-correction of quantum states for quantum computers (which is an interesting topic apparently still in it's infancy). Unlike the study of factoring and discrete logarithms which have undergone lots of study and has a huge knowledge base, we have likely only scratched the surface on other computationally hard algorithms to use as a basis for a public key system. Maybe they aren't really hard at all. At this point, we don't really know. For example, the McElice system's security parameters needed to be updated after some advances in classical information theory knowledge like this...
In contrast, if we knew about this wonderful system but computers to do it aren't big/powerful enough yet to implement it practically, that would be one thing. I think we actually lack the knowledge base to meaningfully change our public key infrastructure today (which should keep all those researcher employed for a good while longer).
I would say that crypto advances and computers go hand in hand...
I would actually say instead that crypto advances go in fits and spirts of understanding instead of tracking the advances of computer technology.
Today, we are apparently not much beyond iterated substitution/permutation block ciphers that have some hardening against linear and differential analysis (things known about by the NSA and applied to the DES algorithm back in the '70s). Nearly all ciphers (and ciphers used in hash algorithms) have been developed along these lines of thinking. When this switches between being an asset and a liablity is anyone's guess, but when it does, there will likely be a fit and spirt of change that goes along with it.
What will be the next big crypto advance is a next generation public key algorithm.
Sadly, that is one area that might be a big crypto advance, but is totally limited by our knowledge** which is again something that has very little to do with the advances in traditional computer technology.
**different computational hard algorithms (e.g., lattice, knapsack, linear error correcting codes, etc) to replace current problems (e.g, factoring and discrete logs, etc), new "trap-door" algorithms that take advantage of those problems with known security parameters and how to develop useful computational asymmetry within those security parameters.
Assuming this policy making board gets what they want... either that the foundation becomes licensed to do business in California as a financial transaction agency (aka a "bribe" as it means the policy board collects some additional revenue) or the foundation is shut down.
What exactly do they think is going to happen if the foundation is shut down? That the software will stop being used or that nobody will be distributing Bitcoin clients? For something published under the GPL?
You can make that argument with just 'bout every one of the professional licensing boards (medical board, bar association, contractor's licensing, etc). The short answer is they don't care what happens, they just want to regulate it. It's basically just a state sponsored guild (not unlike a union).
So, they are a lobbying group that might dogfood to a certain extent. That's like saying that a credit card industry lobby group might need to be registered as a money transmitter because they sometimes pay for things with credit cards.
No, but say if such a lobbying group decided to sponsor/issue a branded credit card to their members as a benefit, they might be walking in the shade of grey...
That's nice. Except for the part where the bitcoin foundation only distributes software.
Except by the own admission, they "standardizes, protects and promotes the use of bitcoin cryptographic money for the benefit of users worldwide." Also, they appear to intend to use foundation money to lobby and fund a group of core developers.
Actions they take when they protect and promote, say if they involve actual sale or exchange of bitcoin (e.g., accepting donations from external developers/lobbyist in bitcoin converting them to USD on their behalf and kicking that money back to fund external developers/lobbyist) may not be looked upon so benignly...
You really based your opinion on the quality of an employer by the fact that a contractor was new, or having an off day?
It appears that your selective reading comprehension of my post was clouded a bit by your zeal to defend your adopted company against any slight with a sad ad-hominem zinger.... Although that was story was tongue in cheek, are you seriously defending the free burrito as a solution to this problem? What if I needed to pick up my kid at daycare? (okay, it was 6:45pm and I made it out by 7:15, and my wife was at home w/ the kids so that wasn't the case). Anyhow, if you re-read, my assessment (before that incident occurred at the end of the day) was that the company was just too big for me.
On the other hand, I seriously doubt this was the first/only time this has **EVER** happened, so my conclusion is that google has no process for this which is a sure sign of a company being simply too big to bother. As a counter example, many moons ago, the company I worked for (which was starting getting a bit too big), was playing parking-lot chicken with the company in the next building in the same office complex. Eventually, they had to hire a valet to pack the cars into the parking lot if you got to the office late you had to valet your car. Although most people retrieved the keys to their car before the switchover to the night-valet lockup (before the valet left, they would unpack the remaining cars so all you had to do was pick up your key and go back to work), old-timers in the company took turns to spam out emails warning there were "N" keys left in the valet "M" minutes before it closed (in case people were eating dinner and forgot, usually someone would walk into the café and yell something after they got that email).
Now, I don't have a solution for something like this for a company the size of google, nor do I think they really care if I could come up with one in 30min in an interview (although I think there's already an app for that), but I digress... Anytime people start dismissing things because it's "not-my-job" (e.g., it's the contractor's fault), it generally indicative of big-company mentality. Some folks enjoy working for big companies, some do not. Variety is the spice of life.
If you only got a free lunch, then you did the interview wrong.
I might have tried that scam a few years ago when I was single since they seem to be spamming me about every 2-3 months. But then I got stranded in the google-plex when they misplaced the keys to my car. Just think what would happen if I let google HR strand me in a foreign country where I don't speak the language...:^(
Besides, I think my wife would kill me if I decided to leave her with the 2 babies while I go off for an interview in a random country that I wanted to visit. She already vetoed a trip to Taiwan to chair a conference session, since she would have to watch the kids by herself and/or she/we would have to suffer/pay for the transpacific flight for the rest of the family...
Interviewed a google a long time ago. I guess I wasn't that serious about wanting to work there, but they emailed me first and hey it's a free lunch in their famous café**
Long story short, I talked to a few interesting folks and none of them asked me any brain teaser crap (which was a pleasant surprise to me), but as I interviewed them (since 1/2 of the interview process trying to figure out if you would actually like the job), I came to the conclusion that the company is just too big.
Funny thing when I was leaving at the end of the day, they gave me a parting 30 minute puzzle to solve which put the nail in the coffin for me.
If the valet at the company you are interviewing misplaces your car key for about 1/2 hour, are they so big they can't even figure out how to keep track of a few car keys, or do they not really care if workers need to leave work at the end of the day (e.g. pick up kids from daycare)? The way the HR person attempted to solve this problem was to offer me a free burrito for my trouble. Sadly, the only answer I had for that problem was to wait it out...
** the café wasn't that great although it's certainly better than the one at the company I current work for...
What excites me about this is that exascale is around what is required to simulate a human brain in its entirety. Who's taking bets on what the first uploaded organism will be?
If we take as historical precident of the human genome, (Craig Ventor followed by James Watson), it will likely be Selmer Bringsford (followed by Gordon Bell, because Seymour Cray was killed in a car crash). Dark horse would be Ray Kurzweil if somehow google beats everyone to the punch...
“I don’t think a major breakthrough is required to get there,” he said. “If we can just keep the funding in place for the next few years, all of this is going to happen. Companies are now selling petawatt lasers commercially, and as we get better at doing this, companies will come into being to make 10 GeV accelerator modules. Then the end users, the chemists and biologists, will come in, and that will lead to more innovations and discoveries.”
1. Start with 1GeV research laser plasma accelerator 2. Demo 2GeV accelerator tied to one of the most powerful petawatt lasers in the world 3. Promise 10GeV if funding continues for next few years... 4. ??? 5. Profit!
Slashdot Mirror
In common usage TSMC is the foundry not Apple (even if apple started buying equipment to process wafers, then it would simply be running a captive fab).
FWIW, the common terms of large foundry contracts have varied greatly over the last few decades. Some examples of "pure" pricing models:
* wafer starts (to reserve some fraction of capacity)
* processed wafers (that had their wafer process monitor circuits working within a range of pre-agreed parameters).
* working die (pre-diced chips that pass a short customer defined wafer probe tests)
Nearly all large contracts are a blend between these pure pricing models and are a function of the times.
When wafers are cheap and factories have lots of capacity (say when a new process comes on line, the older factories may have extra capacity and/or a competitive yield advantage relative to other companies), OR when the process is new and proper yield correlations are unknown, most customers can get the working die pricing model.
When capacity is really tight, a large customer might have to pay some amount of money for a wafer start just to reserve capacity at the factory (even if the processed wafer doesn't yield any working parts). Typically this somewhat negates most of the volume discount they would get and puts the pricing more on par with a smaller customer.
Of course everything is negotiable and there is probably a price for everything. However, given the current pricing environment with the capacity limitations TSMC had with 28nm, I doubt anyone could negotiate pure working die deal now unless they paid a handsome penny for each one.
Most folks in the (fab-less) semiconductor industry are referencing the foundry business model of TSMC. In the historical foundry business model, a foundry is where you can contract to have *custom* metal parts forged. In modern usage, SEH would generally be considered a supplier to TSMC, since the silicon wafer is more of a *standard* part, not a *custom* part.
SEH might have been considered analogous to foundry of sorts since a historical metal-works foundry poured metal into casts and finished them and SEH actually make the silicon ingots and cuts and finishes them into "blank" silicon wafers. However, since it is TSMC which sells custom services, and the wafers are more-or-less standard parts, TSMC are more akin to the foundry provider from a business sense and thus are called that in modern usage.
planes are fly by wire nowadays
Fly by wires isn't the same as fly by wireless which what it would be if you didn't have pilots (unless they went the way of the old "TOW" missals which were connected by fiberoptic cable).
Given the possibilities of jamming and hacking, I think there is less of a chance of a flight deck issue than other ways of hacking/jamming...
The slippery slope is that perfecting this technique is a stepping stone to designer babies.
Apparently, the current technology allows replacement of the whole nucleus, allowing the nucleus of a fertilized egg w/ defective mitochondria to be placed into another fertilized egg with healthy mitochondria. The implication is that this fertilized egg is placed back in the mother (but it could be anyone). Many folks are pretty sure that we are pretty close to the ability to selectively replace a few chromosomes in that nucleus during the swap. Presumably, the first application of this would be to replace/remove a obviously defective chromosome (say trisomy-21 aka down's syndrome) which seems like it might be the first step down the slope.
As mappings of gene-expressions to chromosomes get more accurate, you might imagine that some specific chromosome could be selected for (take all the chromosomes except #k and get that from some other place). Although you might think that the next step might be making a kid with the "best" selection of chromosomes from a single mother and father, identifying a gene in a chromosome is currently a destructive procedure so that's not an easy path to take. Instead, the next logical step (if you allow for N>2 people in the procedure) is to pick a known good chromosome from a 3rd party. With enough 3rd parties involved, you pretty much have a designer baby... ;^)
Of course there is an ethical side to this as well (gotta create/destroy a bunch of *human* fertilized eggs to do this) and that doesn't sit that well with many folks.
there are many, many desperate children that feel so hopeless and lonely right now in some orphanage
Adoption in many countries is very difficult, and plenty of potential parents do not qualify. My wife and I are financially secure, and are very successfully raising two of our own kids. But we had room in our home and our hearts for at least one more, and looked into adoption. We were turned down. The reasons given were that we were too old (I am in my 50s and my wife is in her 40s), and we already have kids of our own, and childless couples would be given priority.
If there really are orphanages full of desperate children, then governments are doing an incredibly poor job of matching them up with willing and capable parents.
Just think of using a donor egg as "adopting" that egg, and I don't think that's as difficult as adopting a baby in many countries (where IVF was available in the first place). The primary thing the 3-person IVF really enables over the commonly available 2-person IVF is that mother can have the option to be biologically related to the child if the mitochondrial dna in her eggs were somehow defective. Another way to look at it, you still have to "adopt" an egg w/ healthy mitochondria from someone (and if you were open to the normal adoption process in the first place, this would appear to be a don't care issue).
Survey says......69!
Hmm, I don't think 69 will bake up a new baby...
Also, mitochonrida don't really have a chromosome...
Im learning assembly language. It will never become obsolete unlike some high level languages.
Nah, go for the metal, I'm learning binary this summer...
0, 1, ok now time for some fun in the sun... ;^)
People are probably beter off reading the wiki...
Key bit of information...
OAM multiplexing can not be implemented in the existing long-haul optical fiber systems, since these systems are based on single-mode fibers, which inherently do not support OAM states of light. Instead, few-mode or multi-mode fibers need to be used. Additional problem for OAM multiplexing implementation is caused by the mode coupling that is present in the fiber, making direct-detection OAM multiplexing still not being realized in long-haul communications. In some specialty fibers, OAM states were transmitted with 97% purity after 20 meters.
Basically this demonstration technique uses specially designed fibers that can carry the "donut" TEM mode required for OAM which is the reason they made a comment that the most likely for fibers the implement this technique "might be to install them to span the short distances between servers on giant 'server farms'"...
I think they just learned the lesson of Las Vegas. Vegas basically started out as an organized money skimming organization fronted by gambling operations (i.e., attempting to make more money by bypassing part the system). After a while, the proprieters found that once they figured out how to run the above board business, it was way more lucrative than the illegal part which means you might as well just concentrate your efforts on the above board business (or suffer the opportunity cost of not doing so).
Basically the lesson is that there's lots of money to be made out in the world by people with skills who get out of the small game and learn how to play the bigger game.
All variance is caused by "malfunction" in the genetic process.
Except of course the variance observed from epigenetic processes (such as DNA methylation which may have a role in cancer and obesity)
But it's not the WOPR! Now THAT would be the bit of movie gear to have! :-)
It would be, but if I recall correctly it was only a prop wired for lights. Still, it would fun to have that in a data center, especially if its one that people tour.
As I recall, that although the WOPR (like other movie computers of that era) was basically a prop with dancing stereo equalizer lights for cinematic effect, I think there was actually an Apple ][ inside controlling an experimental flat panel display on the side.
Although GEB can be an entertaining book, I don't think a course that had GEB as reading material was what the original writer had in mind for a prototypical humanities course***.
*** yes, I know there have been university courses built around GEB, but there have been university philosophy courses built around star trek too
The field of engineering is more like that of a technologist (learning and applying currently well accepted science to perform the job of creating things). A scientific field is more one that is building the base of scientific knowledge which often is incremental from currently well accepted science. Of course there is a big overlap between the two in practice, but when reflected back to undergraduate instruction level, they become suspiciously similar, so much so that it's really hard to distinguish. I would argue that at the undergraduate teaching level, there's not much of a difference between science and technology. All the topics have a "dogma" of some sort which has been decided to be taught by the masters of currucula often in a framework that reflects culture of field. Any so called labs or experiments have well understood answers (and students often are subjected to negative reinforcement if they stray off the beaten path) and are only designed to give a "flavor" of science (like a sad version of a cargo cult).
SO if all students are doing is surveying the history of field (which pretty much sums up undergraduate level instruction), it's pretty much all humanities courses people are taking at that level (since technology is part of the recorded culture of humanity). Probably the only real thing that distinguishs a "humanities" course from a "technology" or "science" course from a knowledge point of view is how recent the history is, what aptitudes/prerequisites are required to understand the material covered, and maybe perhaps how relavent the topic is to future employers...
FWIW, I basically came to this opinion many years ago when I was doing "alumni-calling" for free-pizza and beer. I rang up an alum who was a biologist that graduated 1 year before watson and crick got their nobel prize for discovering DNA. He was still practicing molecular biology, but pretty much had to catch up in grad-school on all stuff that was actually going on during undergrad (as opposed to what the undergrad classes covered). We chatted a while about what he thought about undergraduate level education in biology before this and although he thought he learned a small about the scientific method and the practice of real-world science, the bulk of his classes were pretty much pre-DNA dogma and none of them even mentioned topics around the 3-d structure of molecules or X-ray crystallography which would become his area of specialization (although he did have some exposure to that through talking to a few professors in the chemistry dept.). The undergraduate classes he remembered most, were in fact humanities classes since they had ideas that were more stable over time and he wished he would have taken more of them.
tl;dr
The exception that proves the rule http://en.wikiquote.org/wiki/Wernher_von_Braun
Best summaried by Mr Tom Lehrer...
"Once the rockets are up, who cares where they come down, that's not my department," says Wernher von Braun
Or of course the antectodal fictional situations exist too..
Our studies indicate the weapon is totally useless in warfare.
It's not intended for use in your kind of warfare. It's the perfect peacetime weapon. That's why it's secret.
So it's both immoral *and* unethical?
Yes.
....
Let the engineers figure out a use for it. That's not our concern.
Maybe somebody already has a use for it, one for which it's perfectly designed.
Not that I totally disagree with the indoctrination aspect of many humanities programs, but citing an example about some people trotting out the allegory of Plato's cave isn't really much different than other people trotting out trite observations like Moore's law or Object Oriented Programming or Security through Obscurity because they heard about it somewhere and sadly misapplying them.
I think it's really human nature to be indoctrinated in some way or another as the amount of knowledged to be learned is tremendous and it's really hard to organize in a way that allows for communication or cooperation with others w/o resorting to referencing it to some sort of framework. Unfortunatly, since that framework for understanding is part of the "critical-infrastructure", there's a strong incentive for people with agendas to twist it to their desires. It is after all a human endeavor fraught with human weaknesses.
Centuries ago, the critical-infrastructure of common understanding was probably religion and that was argued by uninformed interpretations until wars broke out. Decades ago it was repeated with VI vs Emacs ;^)
It's not, actually. There are already asymmetric crypto algorithms which are believed to be quantum-resistant. They are typically based on the hardness of vector problems in n-dimensional integer lattices, or problems that have been proven reducible to such problems such as learning with errors.
Interesting read, but it doesn't address the questions at hand:
Are cryptography advances and computer advances hand-in-hand? I believe the answer to that question is still no.
Are big crypto advances in a next generation public key algorithm limited by our knowledge? I believe the answer to that question is still yes.
For example, the McEliece-like crypto system (which seems suspiciously analogus to your LWE paper) did not appear to have any major advances for many years (probably because sending around huge matrices is a real bummer and attempts to reduces this reduced the security parameters). We also have no direct knowledge about its quantum resistance other than some assumptions. Who knows, but its security profile may be very entwined with algorithms for error-correction of quantum states for quantum computers (which is an interesting topic apparently still in it's infancy). Unlike the study of factoring and discrete logarithms which have undergone lots of study and has a huge knowledge base, we have likely only scratched the surface on other computationally hard algorithms to use as a basis for a public key system. Maybe they aren't really hard at all. At this point, we don't really know. For example, the McElice system's security parameters needed to be updated after some advances in classical information theory knowledge like this...
In contrast, if we knew about this wonderful system but computers to do it aren't big/powerful enough yet to implement it practically, that would be one thing. I think we actually lack the knowledge base to meaningfully change our public key infrastructure today (which should keep all those researcher employed for a good while longer).
I would say that crypto advances and computers go hand in hand...
I would actually say instead that crypto advances go in fits and spirts of understanding instead of tracking the advances of computer technology.
Today, we are apparently not much beyond iterated substitution/permutation block ciphers that have some hardening against linear and differential analysis (things known about by the NSA and applied to the DES algorithm back in the '70s). Nearly all ciphers (and ciphers used in hash algorithms) have been developed along these lines of thinking. When this switches between being an asset and a liablity is anyone's guess, but when it does, there will likely be a fit and spirt of change that goes along with it.
What will be the next big crypto advance is a next generation public key algorithm.
Sadly, that is one area that might be a big crypto advance, but is totally limited by our knowledge** which is again something that has very little to do with the advances in traditional computer technology.
**different computational hard algorithms (e.g., lattice, knapsack, linear error correcting codes, etc) to replace current problems (e.g, factoring and discrete logs, etc), new "trap-door" algorithms that take advantage of those problems with known security parameters and how to develop useful computational asymmetry within those security parameters.
Assuming this policy making board gets what they want... either that the foundation becomes licensed to do business in California as a financial transaction agency (aka a "bribe" as it means the policy board collects some additional revenue) or the foundation is shut down.
What exactly do they think is going to happen if the foundation is shut down? That the software will stop being used or that nobody will be distributing Bitcoin clients? For something published under the GPL?
You can make that argument with just 'bout every one of the professional licensing boards (medical board, bar association, contractor's licensing, etc). The short answer is they don't care what happens, they just want to regulate it. It's basically just a state sponsored guild (not unlike a union).
So, they are a lobbying group that might dogfood to a certain extent. That's like saying that a credit card industry lobby group might need to be registered as a money transmitter because they sometimes pay for things with credit cards.
No, but say if such a lobbying group decided to sponsor/issue a branded credit card to their members as a benefit, they might be walking in the shade of grey...
That's nice.
Except for the part where the bitcoin foundation only distributes software.
Except by the own admission, they "standardizes, protects and promotes the use of bitcoin cryptographic money for the
benefit of users worldwide." Also, they appear to intend to use foundation money to lobby and fund a group of core developers.
Actions they take when they protect and promote, say if they involve actual sale or exchange of bitcoin (e.g., accepting donations from external developers/lobbyist in bitcoin converting them to USD on their behalf and kicking that money back to fund external developers/lobbyist) may not be looked upon so benignly...
Of course, right now, it's just a warning...
You really based your opinion on the quality of an employer by the fact that a contractor was new, or having an off day?
It appears that your selective reading comprehension of my post was clouded a bit by your zeal to defend your adopted company against any slight with a sad ad-hominem zinger.... Although that was story was tongue in cheek, are you seriously defending the free burrito as a solution to this problem? What if I needed to pick up my kid at daycare? (okay, it was 6:45pm and I made it out by 7:15, and my wife was at home w/ the kids so that wasn't the case). Anyhow, if you re-read, my assessment (before that incident occurred at the end of the day) was that the company was just too big for me.
On the other hand, I seriously doubt this was the first/only time this has **EVER** happened, so my conclusion is that google has no process for this which is a sure sign of a company being simply too big to bother. As a counter example, many moons ago, the company I worked for (which was starting getting a bit too big), was playing parking-lot chicken with the company in the next building in the same office complex. Eventually, they had to hire a valet to pack the cars into the parking lot if you got to the office late you had to valet your car. Although most people retrieved the keys to their car before the switchover to the night-valet lockup (before the valet left, they would unpack the remaining cars so all you had to do was pick up your key and go back to work), old-timers in the company took turns to spam out emails warning there were "N" keys left in the valet "M" minutes before it closed (in case people were eating dinner and forgot, usually someone would walk into the café and yell something after they got that email).
Now, I don't have a solution for something like this for a company the size of google, nor do I think they really care if I could come up with one in 30min in an interview (although I think there's already an app for that), but I digress... Anytime people start dismissing things because it's "not-my-job" (e.g., it's the contractor's fault), it generally indicative of big-company mentality. Some folks enjoy working for big companies, some do not. Variety is the spice of life.
If you only got a free lunch, then you did the interview wrong.
I might have tried that scam a few years ago when I was single since they seem to be spamming me about every 2-3 months. But then I got stranded in the google-plex when they misplaced the keys to my car. Just think what would happen if I let google HR strand me in a foreign country where I don't speak the language... :^(
Besides, I think my wife would kill me if I decided to leave her with the 2 babies while I go off for an interview in a random country that I wanted to visit. She already vetoed a trip to Taiwan to chair a conference session, since she would have to watch the kids by herself and/or she/we would have to suffer/pay for the transpacific flight for the rest of the family...
Interviewed a google a long time ago. I guess I wasn't that serious about wanting to work there, but they emailed me first and hey it's a free lunch in their famous café**
Long story short, I talked to a few interesting folks and none of them asked me any brain teaser crap (which was a pleasant surprise to me), but as I interviewed them (since 1/2 of the interview process trying to figure out if you would actually like the job), I came to the conclusion that the company is just too big.
Funny thing when I was leaving at the end of the day, they gave me a parting 30 minute puzzle to solve which put the nail in the coffin for me.
If the valet at the company you are interviewing misplaces your car key for about 1/2 hour, are they so big they can't even figure out how to keep track of a few car keys, or do they not really care if workers need to leave work at the end of the day (e.g. pick up kids from daycare)? The way the HR person attempted to solve this problem was to offer me a free burrito for my trouble. Sadly, the only answer I had for that problem was to wait it out...
** the café wasn't that great although it's certainly better than the one at the company I current work for...
What excites me about this is that exascale is around what is required to simulate a human brain in its entirety. Who's taking bets on what the first uploaded organism will be?
If we take as historical precident of the human genome, (Craig Ventor followed by James Watson), it will likely be Selmer Bringsford (followed by Gordon Bell, because Seymour Cray was killed in a car crash). Dark horse would be Ray Kurzweil if somehow google beats everyone to the punch...
“I don’t think a major breakthrough is required to get there,” he said. “If we can just keep the funding in place for the next few years, all of this is going to happen. Companies are now selling petawatt lasers commercially, and as we get better at doing this, companies will come into being to make 10 GeV accelerator modules. Then the end users, the chemists and biologists, will come in, and that will lead to more innovations and discoveries.”
1. Start with 1GeV research laser plasma accelerator
2. Demo 2GeV accelerator tied to one of the most powerful petawatt lasers in the world
3. Promise 10GeV if funding continues for next few years...
4. ???
5. Profit!