I've been studying this (publishing) for some time, in the context of learning, verifying assumptions, and the scientific method.
It turns out that there is really no bar in scientific publishing. It doesn't have to be understandable, nor innovative, nor even correct. You only need to be ethical (ie - don't lie about the data), cite anything that you got from other sources, and show that there is less than a 1-in-20 chance that you are wrong (p > 0.5).
What exactly do we want in a published paper, anyway?
Many cancer studies can't be reproduced. Many studies are statistically significant but valueless (the IQ of people in NYC is higher than Chicago by 1 point: this can be statistically certain but have no practical significance). There are lots and lots of ways to frame the conclusion the wrong way such as confusing correlation with causation, reversed conditionals (if the defendant is innocent, there is a 1 in 1 billion chance that this evidence is wrong), and other logical errors.
Then there's the enormous economic incentive of needing to publish to keep your job, that reviewers will oppose maverick thought and agree with community beliefs, and that no one examines their assumptions.
Would you like to publish a paper? MathGen will write one for you. Pass it around and chances are it will be accepted.
So when I talk to people about my research, the inevitable comment is "you should publish". And my inevitable answer is: why?
What do we want in scientific papers? What are they even for?
The current system apparently jumps to the top of the currently-viewed page when loading is complete. Probably a javascript that does something useful and then jumps as the last action.
It is annoying as hell to be reading somewhere scrolled down in a page and suddenly have the view jump to the top. This happens at the end of the page refresh, and also whenever the refresh timeout happens.
Digging duplicate trenches to lay parallel fiber is wasteful. That's why utilities are "natural monopolies". Getting economic efficiency in such situations usually requires regulation or community ownership.
In the magical land of the oompa-loompas, where Willie Wonka is a benevolent dictator and everything is done for the betterment of their society, this would be important.
Any real issue has arguments both for and against. It's like a mathematical function with many variables, and you have to choose the combination of variables that gives the function the highest value.
In this case the highest value is utility for society, and the variables are the amount of weight you assign to each argument.
Specifically in this case, we assign little weight to "being wasteful because we're digging two trenches" because even though that argument is valid, the utility to society is much lower if we let that consideration drive our choice.
Yeah, I'd *like* to not have to waste effort to have good things, but that's not the world we live in.
Having fiber is more valuable than the expense of digging an extra trench.
The honor code doesn't disallow you from helping someone else through a problem, it disallows you from solving the problem for the other person.
We see this all the time in the discussion forums - someone comes in with "I don't know how to do this", and everyone jumps in to help. If it's a homework problem we can rephrase, use analogy, and solve a similar problem... but we can't outright give the answer or the exact steps for solving.
I asked one of the edX higher-ups (not Dr. Agarwal) about this.
Apparently, there's a lot of interest from companies looking for good talent. For example, people who score in the upper 10% of a high-tech course would be of interest to many companies. Especially in today's market, where putting out a job listing will get thousands of inappropriate resumes.
IIRC, the top 2% of the original AI course (Udacity, not edX) students could optionally have their resume sent to Google for consideration.
The edX higher-up was of the opinion that the future of high-tech hiring would be in the form of online course grades - you would list the courses you took and the grades you got as part of your resume. Companies would advertize for people who had taken specific courses and received certain grades.
There was an episode of MI5 (aired as "Spooks" in the UK) that had this many years ago.
They gave a foreign agent a document to type, and had an eavesdropping device in his office. By recording the keyclicks of the known document, they were able to train the system to decode keyclicks for subsequent documents.
It didn't seem farfetched at the time, it doesn't seem farfetched today.
Your post suggests that research must be done from the benevolent endowment of the government ("grant"), and is the purview of degree'd academics or in pursuit of such a degree.
People speculate that the RDRAND instruction on Ivy Bridge processors has been compromised. If anyone has a spare CPU and motherboard lying around, this can be tested.
The RDRAND internals put the entropy through a random generator before sending the results to the user. This is similar to how rand() works: a single "seed" with limited entropy will generate a long list of seemingly random output, but because there is only one seed the output is predictable and can be reproduced.
To get around this, check the RDRAND data at reset time.
If you had access to a spare CPU and motherboard, you could install your own program in lieu of the BIOS which would catch the RESET vector, get the RDRAND information, initialize a serial port, log the results to a 2nd computer, and force the CPU into RESET.
(For clarity, glossing over some obvious stuff such as storing results in memory and dumping blocks, or dumping to a faster device than a serial port.)
All of the RDRAND tests I've seen have looked at continuously-generated data; which, due to the internal hashing algorithm, would pass even if started with a low-entropy seed. To the best of my knowledge, no one has checked to see if different machines generate the same string of random numbers, or if the starting seed has good entropy.
With a terabyte drive on the logging computer, it should be possible to see if RDRAND has at least 32-bits of entropy: log 4 billion rounds and look for collisions.
RDRAND probably has at least this much entropy, but if not - boy would that paper hit like a bombshell!
The reason for insurance is to spread the risk (and associated costs) over a large population.
When insurance companies pick-and-choose their customers and rates, it invalidates this purpose.
We've been seeing this with health insurance in past decades: ineligible if you have a pre-existing condition, or get dropped if you develop a condition, or get charged more for smoking or being older.
This makes sense from a business perspective, so don't bother saying "what did you expect a business to do?" I'm saying that it makes progressively less sense from the customer's point of view. As these metrics get better, the companies will know exactly how much you will cost them as a customer, and charge the appropriate rates. Why bother with insurance if they know beforehand how much you will need it?
Legally mandated insurance then becomes simple rent-seeking, with no benefit to the consumer.
This particular trend - monitoring the driver's behaviour - is framed as a good idea. Everybody thinks they are a better-than-average driver, so the tradeoff seems like a good deal. You don't care about the big picture because hey! I just saved a bundle on my car insurance!
Here's the big picture: there's no way to verify that the monitoring unit isn't broken, there's no way to verify that the monitoring report is accurate, or that what it's measuring is significant, or that the company isn't skewing the risk. There's no studies that link measured modes with accident risk, no way to tell whether the algorithm for detecting driving modes has flaws, no leeway for corner cases or exceptional conditions, and no way to appeal the decision.
You have a promise from the insurance company that, if you're a safe driver, your rates will go down.
The privacy implications are also important: your driving profile probably tells a great deal about your psychological makeup (how often you use the horn, how sharply you take corners). This would be of enormous benefit to advertisers, profilers, police, and national security agencies. The insurance company can make money by selling this information, but it's OK because it's not financial information.
Ten years from now this will be a problem: insurance companies siphoning money from customers for no benefit.
Perhaps we should be forward-looking this time and prevent useless suffering before it happens.
People only look at their personal short-term gain, with no thought about the long-term consequences. It's paying the Danegeld, nothing less.
I read all the time about this-or-that injustice and oh! the outrage it sparks, but no one wants to do the right thing and fight because it's so hard!
When a cop violates your civil rights, do you take him to court? If no one does, then cops feel free to do whatever they want, and rights violations are everywhere.
When the BSA (business software alliance) demands to search your office without a warrant, when the RIAA offers to settle for less than the court costs, when the border patrol stops and searches your car, or when patent trolls demand license fees, it's all the same: bullies feel free to operate, it's the Danegeld in another form.
If people stood up for their rights and took the bullies to task, there would be a lot less bullying. It would be expensive for the first few people, but in the long run it would be better for everyone. Consider it an investment in your childrens' future: if you fight now, they won't have to fight later.
Next time you read about an injustice, think about what the victim could do to take the bullies to task. Then ask "why didn't they do that?"
Of all the next-generation technologies that we’ve discussed at ET, including carbon nanotubes and graphene, III-V semiconductors that use materials like indium, gallium, and arsenide are by far the most likely to make an a mass market appearance within the next ten years.
[Emphasis mine]
Yeah, that article really seems to know what it's talking about.
An easy fix is to change the "game theory" dynamic.
Currently, we don't pay for usage, we pay for access. The providers get the best value by discouraging use: high monthly fees, data caps, throttling power users, poor facilities, installation fees, and poor customer care.
If the government required providers to charge for usage only, then the providers couldn't increase profits except by increasing use. They would have an incentive to build fast pipes, connect everyone in their area, have customer service that gets people up and running quickly, allow servers, and encourage innovative new applications.
This could be changed without affecting their annual profits - just tally up all the usage in the last year and divide into their current revenues. They would make the same profits next year as last year, but with an to provide better service.
Just another example of how the federal government doesn't really benefit the people.
People always complain that the march of technology will put people out of jobs, historically this has always been proven false.
The issue here is one of similar contexts. If the current situation is the same as all the historical situations, then we can expect the same results. This is simple expectation learning from historical examples.
I claim that the context is different this time. Looking at GDP per capita, we see the aggregate purchasing power is higher than the level needed for everyone to survive.
Standard economic theory builds on the premise of infinite consumption, which comes from the assumptions of infinite population growth and infinite personal consumption.
Infinite population growth is observed to be false. Population is decreasing in most industrialized nations, and US population would be decreasing if it were not for immigration. Even with immigration, the growth rate is almost negative right now and is expected to be negative very soon.
Infinite individual consumption is also observed to be false on the average. When people reach a certain level of comfort, their needs are met and they have no need to consume ever more resources. Individuals won't eat an infinite amount of food if given the chance, won't use an infinite amount of electricity, or buy an infinitely large house. Once their needs are met, consumption levels off.
The upshot is that we have either reached, or are very close to, the level where all production can be satisfied by fewer workers than exist in the population.
This is the difference, this is how the context has changed from the Luddites in the early 1800's. In all previous examples, there was enough demand for production that people could find work in other areas. This time it's different.
We have an ever-growing number of people who are no longer needed to work.
OK, a reasonable point, but what do you suggest? Stop economic progress, the growth in productivity (if we could)? Should we have stopped in 1950? 1920? How would you and I be communicating?
Check it out - it's an easy read, and gives a good introduction from which we can have informed discussion.
To summarize that position, let's take the current economic growth to it's logical conclusion. Imagine a large factory sitting in Arizona which is responsible for producing all consumables needed by the population. It's completely self contained: solar powered, part of its production is diverted to producing replacement solar panels as they wear out. It recycles waste into new products. It's so completely automated that the number of people needed to run the place is negligible.
Everyone in the US is assigned a fixed portion of the factory output in the form of a monthly allowance - say, $1000 worth of production. Each month people order what they need, the factory makes it and has it delivered to the doorstep. 'Sort of like online grocery shopping.
There is no physical reason preventing such a plant; furthermore, there is no physical reason why the plant couldn't divert some of it's production to duplicating itself, so that over time production would double, and then double again.
This is a nice model with no logical inconsistencies that anyone can see, and it's predicted to be the endpoint of our economic development.
No one knows how we transition from the current system to the global factory model yet.
A friend was considering (but ended up not) taking over management of a lithium battery manufacturing plant.
He pointed out that one problem with lithium batteries is heat dissipation.
His plan was to make shaped batteries that could be mounted in the car's unused spaces. Big, flat batteries could be placed in the roof or on the floor, or in the door panels or behind the seats. With a larger surface to volume ration, they would dissipate heat more efficiently than the cylinder form factor. They would also free up space for other purposes.
I never heard anything more on this. Does anyone here more familiar with battery tech know more?
The "ongoing automation of work" has been going on for centuries, and will continue thank goodness.
The farmer has provided food and shelter for all the days of our life, said the turkeys. We therefore conclude that the purpose of a farmer is to provide our food and shelter, and that this will continue for the foreseeable future.
"Predictions of our demise have no merit", they concluded. "We have abundant historical evidence to the contrary."
It's not a "cop out" at all. The party that manages the code doesn't want to remove a feature that there's no logical reason to remove. The petition was one sentence, linked to no debate, made no points and didn't even attempt to negotiate. It could have said, "Do it, because we say so." and it would have been just as informative. I think you need to look up the definition of "cop out", because the petition creators could have actually done something useful, and didn't.
Okay then, lets fix this.
The NSA has compromised products and devices in the design phase - both software and hardware. We don't know which products are compromised or how, but we do know that some are.
Random number generators cannot be verified - it's a computationally infeasible problem. If the NSA has subtly tampered with a product, there's no way to tell from the outside looking in. You *might* be able to tell by looking at the generator source. (Note that the linux random number generator has at least one undocumented source of entropy.)
There is no reasonable way to look at the source code/microcode of the rdrand instruction.
Additionally, there is no way to verify the underlying source of randomness of the rdrand instruction. There could be vulnerabilities on the silicon die.
The whole point of open source is that people can peek at the software and see what's going on.
Since there is no way to inspect the random number generator and no way to verify it's operation, it should not be used by default.
It's a security risk, plain and simple, and risk management should be up to the user. However small the risk is, forcing everyone to take it multiplies the chance that someone will get burned by it.
Here's your logical argument. If Linus wants to debate this, let him address these issues. Linus needs to show the premises wrong, or that the conclusion doesn't follow from the premises.
If he can't, then he should abide by the recommendation.
The NSA has apparently compromised random number hardware and software packages throughout the industry.
Could this be fixed by using an entropy server?
Suppose some group hosted a random number server. A verified source of true randomness which can be trusted by the reputation of the people involved, in the same way that we trust the people who make Tor, Mozilla, and linux.
It would be a single point of failure, but also a single point of defense. We could put all the best practices and best ideas of security into one place, by means of technology, software and legalities. It could be hosted in a privacy-friendly country, it could be monitored and defended by the EFF using legal means, it could use the best technology for generating randomness and have open and easily-inspected software and procedures.
To use the system, a client would:
Generate a public/private key using whatever entropy is on hand
Encrypt the private key using the server's public key and send it to the server
The server returns a packet of random numbers, encrypted using the client's key
The client generates a new key pair using the returned entropy
The client uses that key pair from then on
This is slightly weak because the NSA could record the conversation and "simulate" the client computer to recover the generated keys, but doing this is much harder than cracking weak keys. In the server model the weak key is used once, instead of being used all the time. Also, simulating a computer (including nuances of software version and hardware quirks) is much harder than finding weak keys.
(To find weak keys, gather all the keys you can find and calculate GCD on pairs of keys. In practice, about 1 percent of all keys on the net have common factors. Most of these come from systems with low entropy - headless systems (routers, firewalls, servers) with no user interaction for randomness.)
In one action we could fix the security of much of the software used in the internet.
Any volunteers?
(I'd love to, but it has to be outside the US. I'll donate $1000 towards costs if the idea is viable.)
The Fifth Amendment does not say you can't be compelled to testify... it says you can't be compelled to testify against yourself. If the government grants you immunity, then you cannot -- by definition -- testify "against yourself." A defendant can be compelled to testify, as many have, by granting them immunity.
I agree that a reporter shield law is a good thing, but it is not a constitutional mandate.
Immunity doesn't work that way.
A court doesn't have the authority to grant immunity from a different court, and there are several separate court systems.
For example, a local court can't grant immunity from federal charges, and a federal court can't grant immunity from IRS charges (if your testimony shows that you evaded tax law). The same is true for any agency that's decided that they are the governing legal body for something: FCC, FDA, EPA, NRC - typically none of these agencies is prevented from screwing with you under an immunity agreement.
The best you can get is that the issuing court promises not to pursue charges that it could normally pursue.
Suppose Alice knows that Bob attended a protest. The police would like to arrest everyone at the protest because it wasn't sanctioned - the people didn't apply for a parade permit or permission to gather on public property. Alice believes that once arrested, the police will apply for search warrants to go through Bob's possessions. They won't find anything on the first round, but they will discover something that's both illegal and obscure, so they either argue inevitable discovery or back-fill the information for a new search warrant that turns up the new evidence.
Essentially, Alice wants to prevent government overreach for something that she believes shouldn't be a crime.
The government doesn't act in the interests of the people, and the people have no way to change government. Withholding evidence is a soft way of protesting, one that impedes government overreach without getting you or your friends in trouble.
Just say "I don't remember", and when the police press you, say you "just don't pay attention to these things".
If people could band together and agree to vote out the incumbent (senator, representative, president) whenever one of these incidents crop up, there would be incentive for politicians to better serve the people in order to continue in office. This would mean giving up party loyalty and the idea of "lessor of two evils", which a lot of people won't do. Some congressional elections are quite close, so 2,000 or so petitioners might be enough to swing a future election.
Let your house and senate rep know how you feel about this issue / patriot act and encourage those you know to do the same.
If enough people let their representatives know how they feel obviously those officials who want to be reelected will tend to take notice. We have seen what happens when wikipedia and google go "dark", congressional switchboards melt and the 180's start to pile up.
Fax is considered the best way to contact a congressperson,especially if it is on corporate letterhead.
Suggestion #2:
Take back what's ours through technology and educated practices.
Tor, I2dP and the likes. Let's build a new common internet over the internet. Full strong anonymity and integrity.
Also, Let's go full scale by deploying small wireless routers across the globe creating a real mesh network as internet was designed to be!
Suggestion #4:
What I feel is needed is a true 3rd party, not 3rd, 4th, 5th, and 6th parties, such as Green, Tea Party, Libertarian; we need an agreeable third party that can compete against the two majors without a lot of interference from small parties. We need a consensus third party.
Suggestion #5:
Replace the voting system. Plurality voting will always lead to the mess we have now. The only contribution towards politics I've made in years was to fund Approval Voting video. It's the best compromise for a replacement system. Work to get it allowed at your Town or City level, then we can take it higher.
Suggestion #6:
[Paraphrasing]: Start a social perception that working for evil is evil. Possibly connect this to religious beliefs, but in general shun people who have worked for the system as promoting evil (both in hiring and socially).
The post:
1) this kind of sht is morally wrong
2) thus, working for this kind of sht is morally wrong
3) thus, anybody who works for this kind of sht is going to hell, for whatever your value of 'hell'.
4) you might say that 'i need the money from this gig', but
5) anybody who works for this kind of sht is feeding their kids but is at the same time fscking over the kids' future bigtime. Your kids will not forgive you for being the AC IRL.
From this, it should easily emerge that everybody should just stop working for this sht. No workers, no NSA. There needs to emerge a culture and a movement to encourage it. Shame the spineless coward who works for the Man! Shun him or tell him what he does is evil and his country hates him for it. Spread the word!
As in, if you have a sample rate of 48,000Hz, you can play back a frequency of 24,000Hz (already above the range of human perception). Higher sample rate = more high frequencies you can't hear.
By your logic, one would only need a sample rate of 40,000Hz since human hearing maxes out at 20,000 Hz.
At higher sample rates you get more noise information, which can be used to remove the noise from the signal.
For a much-simplified example, consider sampling at 1 million Hz while looking for a signal at 600 hz. Pick any point in the audio and multiply at this point by 1 wavelength of 600Hz, 600.001Hz, 600.002Hz, and so on for ten frequencies. Average these together and replace in the recorded signal.
The 600 Hz signal will be almost 100% in all 10 samples, but the noise will tend to average out. What gets put back is the original signal with the noise reduced by a factor of 10.
Another way of looking at this is from probabilities. The probability of noise being positive at two consecutive points is 50%. If you have a higher resolution of 100 points between the two original points, then your idea of "average" height of the waveform becomes more accurate.
This glosses over a lot of subtleties and gotchas, but essentially, having a higher sample rate gives you more information about the noise, which can be used to remove the noise from the signal.
Join the class action suit that Rand Paul is bringing against the NSA.
Suggestion #1:
(My idea): If people could band together and agree to vote out the incumbent (senator, representative, president) whenever one of these incidents crop up, there would be incentive for politicians to better serve the people in order to continue in office. This would mean giving up party loyalty and the idea of "lessor of two evils", which a lot of people won't do. Some congressional elections are quite close, so 2,000 or so petitioners might be enough to swing a future election.
Let your house and senate rep know how you feel about this issue / patriot act and encourage those you know to do the same.
If enough people let their representivies know how they feel obviously those officials who want to be reelected will tend to take notice. We have seen what happens when wikipedia and google go "dark", congressional switchboards melt and the 180's start to pile up.
Fax is considered the best way to contact a congressperson,especially if it is on corporate letterhead.
Suggestion #2:
Tor, I2dP and the likes. Let's build a new common internet over the internet. Full strong anonymity and integrity. Transform what an eavesdropper would see in a huge cypherpunk clusterfuck.
Taking back what's ours through technology and educated practices.
Let's go back to the 90' where the internet was a place for knowledgeable and cooperative people.
Someone Added: Let's go full scale by deploying small wireless routers across the globe creating a real mesh network as internet was designed to be!
Suggestion #3:
A first step might be understanding the extent towards which the government actually disagrees with the people. Are we talking about a situation where the government is enacting unpopular policies that people oppose? Or are we talking about a situation where people support the policies? Because the solutions to those two situations are very different.
In many cases involving "national security", I think the situation is closer to the second one. "Tough on X" policies are quite popular, and politicians often pander to people by enacting them. The USA Patriot Act, for example, was hugely popular when it was passed. And in general, politicians get voted out of office more often for being not "tough" on crime and terrorism and whatever else, than for being too over-the-top in pursuing those policies.
Suggestion #4:
What I feel is needed is a true 3rd party, not 3rd, 4th, 5th, and 6th parties, such as Green, Tea Party, Libertarian; we need an agreeable third party that can compete against the two majors without a lot of interference from small parties. We need a consensus third party.
Suggestion #5:
Replace the voting system. Plurality voting will always lead [wikipedia.org] to the mess we have now. The only contribution towards politics I've made in years
Slashdot Mirror
I've been studying this (publishing) for some time, in the context of learning, verifying assumptions, and the scientific method.
It turns out that there is really no bar in scientific publishing. It doesn't have to be understandable, nor innovative, nor even correct. You only need to be ethical (ie - don't lie about the data), cite anything that you got from other sources, and show that there is less than a 1-in-20 chance that you are wrong (p > 0.5).
What exactly do we want in a published paper, anyway?
Many cancer studies can't be reproduced. Many studies are statistically significant but valueless (the IQ of people in NYC is higher than Chicago by 1 point: this can be statistically certain but have no practical significance). There are lots and lots of ways to frame the conclusion the wrong way such as confusing correlation with causation, reversed conditionals (if the defendant is innocent, there is a 1 in 1 billion chance that this evidence is wrong), and other logical errors.
Then there's the enormous economic incentive of needing to publish to keep your job, that reviewers will oppose maverick thought and agree with community beliefs, and that no one examines their assumptions.
Would you like to publish a paper? MathGen will write one for you. Pass it around and chances are it will be accepted.
So when I talk to people about my research, the inevitable comment is "you should publish". And my inevitable answer is: why?
What do we want in scientific papers? What are they even for?
The current system apparently jumps to the top of the currently-viewed page when loading is complete. Probably a javascript that does something useful and then jumps as the last action.
It is annoying as hell to be reading somewhere scrolled down in a page and suddenly have the view jump to the top. This happens at the end of the page refresh, and also whenever the refresh timeout happens.
Please fix this - just get rid of the jump.
Digging duplicate trenches to lay parallel fiber is wasteful. That's why utilities are "natural monopolies". Getting economic efficiency in such situations usually requires regulation or community ownership.
In the magical land of the oompa-loompas, where Willie Wonka is a benevolent dictator and everything is done for the betterment of their society, this would be important.
Any real issue has arguments both for and against. It's like a mathematical function with many variables, and you have to choose the combination of variables that gives the function the highest value.
In this case the highest value is utility for society, and the variables are the amount of weight you assign to each argument.
Specifically in this case, we assign little weight to "being wasteful because we're digging two trenches" because even though that argument is valid, the utility to society is much lower if we let that consideration drive our choice.
Yeah, I'd *like* to not have to waste effort to have good things, but that's not the world we live in.
Having fiber is more valuable than the expense of digging an extra trench.
I've been in several online courses.
The honor code doesn't disallow you from helping someone else through a problem, it disallows you from solving the problem for the other person.
We see this all the time in the discussion forums - someone comes in with "I don't know how to do this", and everyone jumps in to help. If it's a homework problem we can rephrase, use analogy, and solve a similar problem... but we can't outright give the answer or the exact steps for solving.
I asked one of the edX higher-ups (not Dr. Agarwal) about this.
Apparently, there's a lot of interest from companies looking for good talent. For example, people who score in the upper 10% of a high-tech course would be of interest to many companies. Especially in today's market, where putting out a job listing will get thousands of inappropriate resumes.
IIRC, the top 2% of the original AI course (Udacity, not edX) students could optionally have their resume sent to Google for consideration.
The edX higher-up was of the opinion that the future of high-tech hiring would be in the form of online course grades - you would list the courses you took and the grades you got as part of your resume. Companies would advertize for people who had taken specific courses and received certain grades.
There was an episode of MI5 (aired as "Spooks" in the UK) that had this many years ago.
They gave a foreign agent a document to type, and had an eavesdropping device in his office. By recording the keyclicks of the known document, they were able to train the system to decode keyclicks for subsequent documents.
It didn't seem farfetched at the time, it doesn't seem farfetched today.
Sounds like a thesis. Go get a grant.
Interesting.
Your post suggests that research must be done from the benevolent endowment of the government ("grant"), and is the purview of degree'd academics or in pursuit of such a degree.
Are there no Gentleman Scientists any more?
I suppose in today's terms we would call them Makers. Moxie Marlinspike probably isn't a credible researcher.
People speculate that the RDRAND instruction on Ivy Bridge processors has been compromised. If anyone has a spare CPU and motherboard lying around, this can be tested.
The RDRAND internals put the entropy through a random generator before sending the results to the user. This is similar to how rand() works: a single "seed" with limited entropy will generate a long list of seemingly random output, but because there is only one seed the output is predictable and can be reproduced.
To get around this, check the RDRAND data at reset time.
If you had access to a spare CPU and motherboard, you could install your own program in lieu of the BIOS which would catch the RESET vector, get the RDRAND information, initialize a serial port, log the results to a 2nd computer, and force the CPU into RESET.
(For clarity, glossing over some obvious stuff such as storing results in memory and dumping blocks, or dumping to a faster device than a serial port.)
All of the RDRAND tests I've seen have looked at continuously-generated data; which, due to the internal hashing algorithm, would pass even if started with a low-entropy seed. To the best of my knowledge, no one has checked to see if different machines generate the same string of random numbers, or if the starting seed has good entropy.
With a terabyte drive on the logging computer, it should be possible to see if RDRAND has at least 32-bits of entropy: log 4 billion rounds and look for collisions.
RDRAND probably has at least this much entropy, but if not - boy would that paper hit like a bombshell!
The reason for insurance is to spread the risk (and associated costs) over a large population.
When insurance companies pick-and-choose their customers and rates, it invalidates this purpose.
We've been seeing this with health insurance in past decades: ineligible if you have a pre-existing condition, or get dropped if you develop a condition, or get charged more for smoking or being older.
This makes sense from a business perspective, so don't bother saying "what did you expect a business to do?" I'm saying that it makes progressively less sense from the customer's point of view. As these metrics get better, the companies will know exactly how much you will cost them as a customer, and charge the appropriate rates. Why bother with insurance if they know beforehand how much you will need it?
Legally mandated insurance then becomes simple rent-seeking, with no benefit to the consumer.
This particular trend - monitoring the driver's behaviour - is framed as a good idea. Everybody thinks they are a better-than-average driver, so the tradeoff seems like a good deal. You don't care about the big picture because hey! I just saved a bundle on my car insurance!
Here's the big picture: there's no way to verify that the monitoring unit isn't broken, there's no way to verify that the monitoring report is accurate, or that what it's measuring is significant, or that the company isn't skewing the risk. There's no studies that link measured modes with accident risk, no way to tell whether the algorithm for detecting driving modes has flaws, no leeway for corner cases or exceptional conditions, and no way to appeal the decision.
You have a promise from the insurance company that, if you're a safe driver, your rates will go down.
The privacy implications are also important: your driving profile probably tells a great deal about your psychological makeup (how often you use the horn, how sharply you take corners). This would be of enormous benefit to advertisers, profilers, police, and national security agencies. The insurance company can make money by selling this information, but it's OK because it's not financial information.
Ten years from now this will be a problem: insurance companies siphoning money from customers for no benefit.
Perhaps we should be forward-looking this time and prevent useless suffering before it happens.
This is an example of long-term thinking.
People only look at their personal short-term gain, with no thought about the long-term consequences. It's paying the Danegeld, nothing less.
I read all the time about this-or-that injustice and oh! the outrage it sparks, but no one wants to do the right thing and fight because it's so hard!
When a cop violates your civil rights, do you take him to court? If no one does, then cops feel free to do whatever they want, and rights violations are everywhere.
When the BSA (business software alliance) demands to search your office without a warrant, when the RIAA offers to settle for less than the court costs, when the border patrol stops and searches your car, or when patent trolls demand license fees, it's all the same: bullies feel free to operate, it's the Danegeld in another form.
If people stood up for their rights and took the bullies to task, there would be a lot less bullying. It would be expensive for the first few people, but in the long run it would be better for everyone. Consider it an investment in your childrens' future: if you fight now, they won't have to fight later.
Next time you read about an injustice, think about what the victim could do to take the bullies to task. Then ask "why didn't they do that?"
Disqus comments are blocked by ghostery. Are you saying that people shouldn't run ghostery?
Instead of being all high and mighty, perhaps next time you should help out...
Why?
Historically speaking, helping out doesn't help. Putting a warning (of sorts) in this forum lets a whole bunch of potential readers save time.
That's not being "high and mighty", that's making an assessment of quality... which is what our teachers do and what we are taught to do ourselves.
"High and mighty" is calling your own actions better than the actions of someone else, which is what you did.
Of all the next-generation technologies that we’ve discussed at ET, including carbon nanotubes and graphene, III-V semiconductors that use materials like indium, gallium, and arsenide are by far the most likely to make an a mass market appearance within the next ten years.
[Emphasis mine]
Yeah, that article really seems to know what it's talking about.
An easy fix is to change the "game theory" dynamic.
Currently, we don't pay for usage, we pay for access. The providers get the best value by discouraging use: high monthly fees, data caps, throttling power users, poor facilities, installation fees, and poor customer care.
If the government required providers to charge for usage only, then the providers couldn't increase profits except by increasing use. They would have an incentive to build fast pipes, connect everyone in their area, have customer service that gets people up and running quickly, allow servers, and encourage innovative new applications.
This could be changed without affecting their annual profits - just tally up all the usage in the last year and divide into their current revenues. They would make the same profits next year as last year, but with an to provide better service.
Just another example of how the federal government doesn't really benefit the people.
I'm confused.
Were the corporations expected to challenge this? Is that how the system works?
I thought the courts enforced the law, by disallowing blatantly illegal procedures.
Does this mean that anything not specifically challenged is OK?
People always complain that the march of technology will put people out of jobs, historically this has always been proven false.
The issue here is one of similar contexts. If the current situation is the same as all the historical situations, then we can expect the same results. This is simple expectation learning from historical examples.
I claim that the context is different this time. Looking at GDP per capita, we see the aggregate purchasing power is higher than the level needed for everyone to survive.
Standard economic theory builds on the premise of infinite consumption, which comes from the assumptions of infinite population growth and infinite personal consumption.
Infinite population growth is observed to be false. Population is decreasing in most industrialized nations, and US population would be decreasing if it were not for immigration. Even with immigration, the growth rate is almost negative right now and is expected to be negative very soon.
Infinite individual consumption is also observed to be false on the average. When people reach a certain level of comfort, their needs are met and they have no need to consume ever more resources. Individuals won't eat an infinite amount of food if given the chance, won't use an infinite amount of electricity, or buy an infinitely large house. Once their needs are met, consumption levels off.
The upshot is that we have either reached, or are very close to, the level where all production can be satisfied by fewer workers than exist in the population.
This is the difference, this is how the context has changed from the Luddites in the early 1800's. In all previous examples, there was enough demand for production that people could find work in other areas. This time it's different.
We have an ever-growing number of people who are no longer needed to work.
OK, a reasonable point, but what do you suggest? Stop economic progress, the growth in productivity (if we could)? Should we have stopped in 1950? 1920? How would you and I be communicating?
Here's a well-written example of a possible alternate system.
Check it out - it's an easy read, and gives a good introduction from which we can have informed discussion.
To summarize that position, let's take the current economic growth to it's logical conclusion. Imagine a large factory sitting in Arizona which is responsible for producing all consumables needed by the population. It's completely self contained: solar powered, part of its production is diverted to producing replacement solar panels as they wear out. It recycles waste into new products. It's so completely automated that the number of people needed to run the place is negligible.
Everyone in the US is assigned a fixed portion of the factory output in the form of a monthly allowance - say, $1000 worth of production. Each month people order what they need, the factory makes it and has it delivered to the doorstep. 'Sort of like online grocery shopping.
There is no physical reason preventing such a plant; furthermore, there is no physical reason why the plant couldn't divert some of it's production to duplicating itself, so that over time production would double, and then double again.
This is a nice model with no logical inconsistencies that anyone can see, and it's predicted to be the endpoint of our economic development.
No one knows how we transition from the current system to the global factory model yet.
A friend was considering (but ended up not) taking over management of a lithium battery manufacturing plant.
He pointed out that one problem with lithium batteries is heat dissipation.
His plan was to make shaped batteries that could be mounted in the car's unused spaces. Big, flat batteries could be placed in the roof or on the floor, or in the door panels or behind the seats. With a larger surface to volume ration, they would dissipate heat more efficiently than the cylinder form factor. They would also free up space for other purposes.
I never heard anything more on this. Does anyone here more familiar with battery tech know more?
The "ongoing automation of work" has been going on for centuries, and will continue thank goodness.
The farmer has provided food and shelter for all the days of our life, said the turkeys. We therefore conclude that the purpose of a farmer is to provide our food and shelter, and that this will continue for the foreseeable future.
"Predictions of our demise have no merit", they concluded. "We have abundant historical evidence to the contrary."
Linux uses the Ivy Bridge random number generator in the kernel, along with other sources of randomness.
That makes it OK, because as everyone knows, mixing the other sources with a predictable string makes the output even more random!
Didn't Linus completely settle this issue?
It's not a "cop out" at all. The party that manages the code doesn't want to remove a feature that there's no logical reason to remove. The petition was one sentence, linked to no debate, made no points and didn't even attempt to negotiate. It could have said, "Do it, because we say so." and it would have been just as informative. I think you need to look up the definition of "cop out", because the petition creators could have actually done something useful, and didn't.
Okay then, lets fix this.
The NSA has compromised products and devices in the design phase - both software and hardware. We don't know which products are compromised or how, but we do know that some are.
Random number generators cannot be verified - it's a computationally infeasible problem. If the NSA has subtly tampered with a product, there's no way to tell from the outside looking in. You *might* be able to tell by looking at the generator source. (Note that the linux random number generator has at least one undocumented source of entropy.)
There is no reasonable way to look at the source code/microcode of the rdrand instruction.
Additionally, there is no way to verify the underlying source of randomness of the rdrand instruction. There could be vulnerabilities on the silicon die.
The whole point of open source is that people can peek at the software and see what's going on.
Since there is no way to inspect the random number generator and no way to verify it's operation, it should not be used by default.
It's a security risk, plain and simple, and risk management should be up to the user. However small the risk is, forcing everyone to take it multiplies the chance that someone will get burned by it.
Here's your logical argument. If Linus wants to debate this, let him address these issues. Linus needs to show the premises wrong, or that the conclusion doesn't follow from the premises.
If he can't, then he should abide by the recommendation.
The NSA has apparently compromised random number hardware and software packages throughout the industry.
Could this be fixed by using an entropy server?
Suppose some group hosted a random number server. A verified source of true randomness which can be trusted by the reputation of the people involved, in the same way that we trust the people who make Tor, Mozilla, and linux.
It would be a single point of failure, but also a single point of defense. We could put all the best practices and best ideas of security into one place, by means of technology, software and legalities. It could be hosted in a privacy-friendly country, it could be monitored and defended by the EFF using legal means, it could use the best technology for generating randomness and have open and easily-inspected software and procedures.
To use the system, a client would:
This is slightly weak because the NSA could record the conversation and "simulate" the client computer to recover the generated keys, but doing this is much harder than cracking weak keys. In the server model the weak key is used once, instead of being used all the time. Also, simulating a computer (including nuances of software version and hardware quirks) is much harder than finding weak keys.
(To find weak keys, gather all the keys you can find and calculate GCD on pairs of keys. In practice, about 1 percent of all keys on the net have common factors. Most of these come from systems with low entropy - headless systems (routers, firewalls, servers) with no user interaction for randomness.)
In one action we could fix the security of much of the software used in the internet.
Any volunteers?
(I'd love to, but it has to be outside the US. I'll donate $1000 towards costs if the idea is viable.)
The Fifth Amendment does not say you can't be compelled to testify... it says you can't be compelled to testify against yourself. If the government grants you immunity, then you cannot -- by definition -- testify "against yourself." A defendant can be compelled to testify, as many have, by granting them immunity.
I agree that a reporter shield law is a good thing, but it is not a constitutional mandate.
Immunity doesn't work that way.
A court doesn't have the authority to grant immunity from a different court, and there are several separate court systems.
For example, a local court can't grant immunity from federal charges, and a federal court can't grant immunity from IRS charges (if your testimony shows that you evaded tax law). The same is true for any agency that's decided that they are the governing legal body for something: FCC, FDA, EPA, NRC - typically none of these agencies is prevented from screwing with you under an immunity agreement.
The best you can get is that the issuing court promises not to pursue charges that it could normally pursue.
Immunity is really a very narrow protection.
How about protesting against unfair laws?
Suppose Alice knows that Bob attended a protest. The police would like to arrest everyone at the protest because it wasn't sanctioned - the people didn't apply for a parade permit or permission to gather on public property. Alice believes that once arrested, the police will apply for search warrants to go through Bob's possessions. They won't find anything on the first round, but they will discover something that's both illegal and obscure, so they either argue inevitable discovery or back-fill the information for a new search warrant that turns up the new evidence.
Essentially, Alice wants to prevent government overreach for something that she believes shouldn't be a crime.
The government doesn't act in the interests of the people, and the people have no way to change government. Withholding evidence is a soft way of protesting, one that impedes government overreach without getting you or your friends in trouble.
Just say "I don't remember", and when the police press you, say you "just don't pay attention to these things".
From a previous post, here's the collected list of suggested actions people can take to help change the situation.
Have more ideas? Reply below & I'll add them to the list
Links worthy of attention:
Suggestion #1:
If people could band together and agree to vote out the incumbent (senator, representative, president) whenever one of these incidents crop up, there would be incentive for politicians to better serve the people in order to continue in office. This would mean giving up party loyalty and the idea of "lessor of two evils", which a lot of people won't do. Some congressional elections are quite close, so 2,000 or so petitioners might be enough to swing a future election.
Let your house and senate rep know how you feel about this issue / patriot act and encourage those you know to do the same.
If enough people let their representatives know how they feel obviously those officials who want to be reelected will tend to take notice. We have seen what happens when wikipedia and google go "dark", congressional switchboards melt and the 180's start to pile up.
Fax is considered the best way to contact a congressperson,especially if it is on corporate letterhead.
Suggestion #2:
Take back what's ours through technology and educated practices.
Tor, I2dP and the likes. Let's build a new common internet over the internet. Full strong anonymity and integrity.
Also, Let's go full scale by deploying small wireless routers across the globe creating a real mesh network as internet was designed to be!
Suggestion #4:
What I feel is needed is a true 3rd party, not 3rd, 4th, 5th, and 6th parties, such as Green, Tea Party, Libertarian; we need an agreeable third party that can compete against the two majors without a lot of interference from small parties. We need a consensus third party.
Suggestion #5:
Replace the voting system. Plurality voting will always lead to the mess we have now. The only contribution towards politics I've made in years was to fund Approval Voting video. It's the best compromise for a replacement system. Work to get it allowed at your Town or City level, then we can take it higher.
Suggestion #6:
[Paraphrasing]: Start a social perception that working for evil is evil. Possibly connect this to religious beliefs, but in general shun people who have worked for the system as promoting evil (both in hiring and socially).
The post:
1) this kind of sht is morally wrong
2) thus, working for this kind of sht is morally wrong
3) thus, anybody who works for this kind of sht is going to hell, for
whatever your value of 'hell'.
4) you might say that 'i need the money from this gig', but
5) anybody who works for this kind of sht is feeding their kids but is
at the same time fscking over the kids' future bigtime. Your kids will
not forgive you for being the AC IRL.
From this, it should easily emerge that everybody should just stop working for this sht. No workers, no NSA. There needs to emerge a culture and a movement to encourage it. Shame the spineless coward who works for the Man! Shun him or tell him what he does is evil and his country hates him for it. Spread the word!
As in, if you have a sample rate of 48,000Hz, you can play back a frequency of 24,000Hz (already above the range of human perception). Higher sample rate = more high frequencies you can't hear.
By your logic, one would only need a sample rate of 40,000Hz since human hearing maxes out at 20,000 Hz.
At higher sample rates you get more noise information, which can be used to remove the noise from the signal.
For a much-simplified example, consider sampling at 1 million Hz while looking for a signal at 600 hz. Pick any point in the audio and multiply at this point by 1 wavelength of 600Hz, 600.001Hz, 600.002Hz, and so on for ten frequencies. Average these together and replace in the recorded signal.
The 600 Hz signal will be almost 100% in all 10 samples, but the noise will tend to average out. What gets put back is the original signal with the noise reduced by a factor of 10.
Another way of looking at this is from probabilities. The probability of noise being positive at two consecutive points is 50%. If you have a higher resolution of 100 points between the two original points, then your idea of "average" height of the waveform becomes more accurate.
This glosses over a lot of subtleties and gotchas, but essentially, having a higher sample rate gives you more information about the noise, which can be used to remove the noise from the signal.
From a previous post, here's the collected list of suggested actions people can take to help change the situation.
Have more ideas? Please post below.
Links worthy of attention:
http://anticorruptionact.org/ [anticorruptionact.org]
http://www.ted.com/talks/lawrence_lessig_we_the_people_and_the_republic_we_must_reclaim.html [ted.com]
http://action.fairelectionsnow.org/fairelections [fairelectionsnow.org]
http://represent.us/ [represent.us]
http://www.protectourdemocracy.com/ [protectourdemocracy.com]
http://www.wolf-pac.com/ [wolf-pac.com]
https://www.unpac.org/ [unpac.org]
http://www.thirty-thousand.org/ [thirty-thousand.org]
Join the class action suit that Rand Paul is bringing against the NSA.
Suggestion #1:
(My idea): If people could band together and agree to vote out the incumbent (senator, representative, president) whenever one of these incidents crop up, there would be incentive for politicians to better serve the people in order to continue in office. This would mean giving up party loyalty and the idea of "lessor of two evils", which a lot of people won't do. Some congressional elections are quite close, so 2,000 or so petitioners might be enough to swing a future election.
Let your house and senate rep know how you feel about this issue / patriot act and encourage those you know to do the same.
If enough people let their representivies know how they feel obviously those officials who want to be reelected will tend to take notice. We have seen what happens when wikipedia and google go "dark", congressional switchboards melt and the 180's start to pile up.
Fax is considered the best way to contact a congressperson,especially if it is on corporate letterhead.
Suggestion #2:
Tor, I2dP and the likes. Let's build a new common internet over the internet. Full strong anonymity and integrity. Transform what an
eavesdropper would see in a huge cypherpunk clusterfuck.
Taking back what's ours through technology and educated practices.
Let's go back to the 90' where the internet was a place for knowledgeable and cooperative people.
Someone Added: Let's go full scale by deploying small wireless routers across the globe creating a real mesh network as internet was designed to be!
Suggestion #3:
A first step might be understanding the extent towards which the government actually disagrees with the people. Are we talking about a situation where the government is enacting unpopular policies that people oppose? Or are we talking about a situation where people support the policies? Because the solutions to those two situations are very different.
In many cases involving "national security", I think the situation is closer to the second one. "Tough on X" policies are quite popular, and politicians often pander to people by enacting them. The USA Patriot Act, for example, was hugely popular when it was passed. And in general, politicians get voted out of office more often for being not "tough" on crime and terrorism and whatever else, than for being too over-the-top in pursuing those policies.
Suggestion #4:
What I feel is needed is a true 3rd party, not 3rd, 4th, 5th, and 6th parties, such as Green, Tea Party, Libertarian; we need an agreeable third party that can compete against the two majors without a lot of interference from small parties. We need a consensus third party.
Suggestion #5:
Replace the voting system. Plurality voting will always lead [wikipedia.org] to the mess we have now. The only contribution towards politics I've made in years