The great majority of vaccines are extremely valuable. Unfortunately the same can't be said in general for everything provided by the medical industry, and uninformed / uneducated people may not understand the fundamental difference between say vaccines and over-prescribed pain killers.
Its easy to think that everyone should be informed, but despite our best efforts one out of ten of the population is in the bottom 10% in terms of understanding things. That is still a lot of people and we need to help them know how to decide.
I ran into a social media post recently where a mother was saying that she didn't trust doctors do inject "chemicals" into her children, and showed a (true) story about a child who died of a flu vaccine.
The people opposing vaccination are not evil, they just have not be taught statistics and rational thought. They have no idea how vaccines work. They don't know what sources of information to trust.
Plastics containing just H and C can probably be burned without much processing. The problem is that most recycled plastic is a mix of all sorts, including a lot of popular Cl containing plastics that are really nasty to burn.
If there are sources of sufficiently pure hydrocarbon plastics (polyethylene, polypropylene and the like) and if this is more cost effective and energy effective than other methods, its fine. I expect the bigger problem is the initial purification. Maybe there is a solution to separating out the other plastics?
Sometimes there isn't a clear boarder between sensitive and non-sensitive information. Many people do work at home, or on personal laptops while traveling. While that certainly woudln't include classified information, it might be related to work that is sensitive - sometimes just in work emails.
Often this work is done on people's personal time, so expecting them to go to extra effort to carry additional devices is likely to result in them just not doing the work, and a reduction in productivity.
If I were required to carry a work laptop when on personal travel, I would stop doing work for free when traveling.
Its easy to say "we want everyone to have their personal star-ship by 2030", but if you don't have a plan to get there you just look foolish when you fail.
If there is a general plan, then they should show it. At what rate does solar and wind production need to be ramped up. That tells you about how many factories to build, how many workers etc. Large projects know how to do this.
If it needs new technology then say that: "we've calculated that we can ramp up solar and wind quickly enough but will need *new technology* for energy storage". That tells people what is missing and where to put R&D.
Otherwise, why 10 years? Why not 5, or 1, or tomorrow? What is the argument that 10 years is the right time scale. To me it seems absurdly short - they sent 10 years building a single railway overpass near my house, and 30 rebuilding a single damaged bridge. How can anyone imagine a huge change in US infrastructure in 10 years?
If there is a plan, then lets see it. If not, they are just discrediting legitimate programs to reduce CO2 emissions .
The problem is the crime predicting software may cause police to investigate people against whom there is no solid evidence. Investigation is not harmless -it can cause significant inconvenience and in some cases physical risk. It can also uncover minor crimes that are committed by many people, but usually not noticed. So harm can be caused to people without anything like due process.
Its very difficult to keep bias out of machine learning data sets. If past bias has caused an unusually high number of arrests for some minority group, information that is a proxy for membership in that group will cause more investigations, which will lead to a higher rate of detected crimes and a positive feedback. Its exactly the same sort of feedback that operates with humans, but without the higher reasoning functions that allow some humans to counteract this pattern matching.
Self flying / taxxing aircraft aren't that difficult a problem and they would interact well with automated air traffic control. As long as there are humans though, I think fully automated traffic control would be difficult. Instructions are still given by voice and they can be misunderstood and / or garbled etc.
Automation is also very poor at dealing with unusual situations - say a major earthquake at SFO that might have damaged runways, or a piece of debris falling off of a plane and creating a hazard, or a confused inexperienced (or sometimes experienced) pilot blundering into the wrong airspace. I think some humans will be needed to deal with these sorts of situations.
A company wants verifiable identities on the people who use their site, which will increase the value of the data that company sells to their customers.
Convincing the users (product) to go along is just marketing.
That is true only if you are assuming its the earth that is eclipsing the moon. Other options are however substantially less likely, and less desirable.
If it works as described, and if the cost is reasonable, this would be fantastic for aircraft ice protection. Valuable for small planes that don't have de-ice heat, and for airliners to need less area covered with de-ice gear.
If it works. This type of success has been described before, so I'll wait a bit before I believe it. If it does really work though, its great .
Computers are very fast in terms of quantitative performance metrics, but I think are less capable of doing useful things that we expected 40 years ago. Probably that is because things like AI turned out to be far harder than most people expected.
Can I have intelligent conversations with my computer: no (no HAL 9000)
Are there regular commercial flights to the moon, and a permanent base on Mars? no. I can't even fly in a supersonic passenger plane anymore.
Fusion reactors? No. Flying cars? No. Undersea cites: No. Humanoid robots helping me round the house: no.
We have fallen far short of what was predicted when I was young. Of course that may be because the predictions were crazy, not because we have in some way failed.
My feeling is that we failed at space. The other things turned out to be harder than we expected.
There are so many attacks on privacy that the public can't resist them all. Its easy to suggest that people not purchase products that spy on them, but when it becomes common for an entire industry (cell phones, TVs, etc), eventually the consumers just give up.
I think we are heading toward 24/7 surveillance that is almost impossible to avoid. That date will then be sold to the highest bidder - or hacked. Companies will use machine learning to look for marketing opportunities. Governments will do the same looking for signs of disloyalty.
Its a rather scary future, but very difficult to avoid.
To get rid of the panopticon, we need to find some other source of revenue. At the moment there is no way to guarantee that if you pay for services you will no longer be tracked, so we are in a sort of latch-up.
People understand this as individuals - they run marathons, climb mountains, learn all sort of new "useless" skills, but we've forgotten that its important to do as a society as well. We are so focused on "efficiency" that we forget to have goals
Groups with obvious CP related names? Maybe, but it could also be an organization that wants to outlaw strong encryption so that they can monitor all communication, and which is safe from prosecution under CP laws.
There are many claims (which of course may be false) that the FBI is the worlds largest distributor of child porn - done of course to attempt to catch (entrap?) peole trading illegal material.
This may legitimately be a bunch of incredibly stupid child pornographers, in which case I hope they spend a good long time in prison. OTHO, its sure convenient for organizations that want to stop encryption.
If these problems apply to payroll and purchasing systems, then its a problem that should be fixed. If they apply to actually missile systems, then of course its a whole different kettle of fish.
I hope and assume any missile systems or classified systems are air-gapped, and things like 2-factor authentication and anti-virus do not apply. Security is guys with guns who shoot anyone who crosses the air-gap without authorization.
I also hope that any report on vulnerabilities of missile systems would be classified.
I'd assume that a machine CBP touched should be immediately destroyed. I have no faith that they haven't installed something or that they may have made security mistakes that allowed the machine to be compromised. I'd probably tell them to keep it.
Yes, its a remarkable technology. You can close the door and avoid distractions.... or.... I know this is a leap.... open the door and collaborate. You can do *both*.
I've never understood why companies pay workers a ton of money, then put them in situations where they can't work efficiently. Then Apple goes and builds a zillion-dollar headquarters that STILL doesn't provide efficient working spaces.
I'd like to believe that high level managers are just clueless an think that if a bunch of smart engineers talk with each other, flying cars, robots and warp drive spaceships will just appear. I doubt it though - I think its just elitism - an office becomes a mark of status to separate high level managers from the workers.
It would be very difficult to do a valid study. There is a lot of difference between lying about a circle or cross on hidden playing card, and lying about whether you gave the Russians your password, which is again different from whether you cheated on your wife.
In many case truth vs lie may not even be all that well defined.
For those and many other reasons, until I see refereed journal studies, I have 0 confidence in any "lie detector".
At some point (not far away these days) the cooling system starts to become as large as the chips it is cooling, and then there isn't a lot of benefit for higher power densities.
I think the energy per gate switch has to come down. There is a fundamental (I think) limit of several times thermal energy (~1/40ev at room temperature). If you cool below room temperature the Carnot efficiency of the refrigerator keeps you from going much further.
I don't know how close current devices are to KT energy these days.
In a field where personal opinions are as important as they are in HR, its going to be difficult to find an unbiased training set. The resulting AI could easily make strongly biased decisions.
Clearly this is an utterly insane idea if there is any chance of unauthorized parties taking over cars and say driving them into rivers, or each other by the tens of thousands or millions. Talk about a terrorists dream.
So I assume the there is some perfect security solution that makes this all work. Tesla and the like must be much better at security than say the NSA - which got itself badly hacked a while ago.
One problem that arises is that science puts a lot of weight on research published in high impact refereed journals. There is a general believe that a paper in Nature or Science or the like has been carefully reviewed. The really is due to the reputation of those journals.
Open journals may eventually gain the same reputation, but it is not an instant process. There are a lot of junk journals out there that will claim to referee, but are really "pay to publish" - anything you send them will be published.
So if science moves away from traditional journals, how can the interested, but non-expert (in the field) public know what has actually been reviewed. How can funding agencies judge the quality of research form publicly funded science?
I expect that in the long term some open journals will develop widely recognized reputations. (remember a reputation just in a field doesn't help with the above problems). I just don't see how to get there quickly.
Is it plausibly useful for very high altitude drones, mars aircraft and the like? What is the effective exhaust velocity? eg is there any regime where it is more efficient than an electric motor and propeller?
Still its a cute concept, even if it isn't practical.
Slashdot Mirror
The great majority of vaccines are extremely valuable. Unfortunately the same can't be said in general for everything provided by the medical industry, and uninformed / uneducated people may not understand the fundamental difference between say vaccines and over-prescribed pain killers.
Its easy to think that everyone should be informed, but despite our best efforts one out of ten of the population is in the bottom 10% in terms of understanding things. That is still a lot of people and we need to help them know how to decide.
I ran into a social media post recently where a mother was saying that she didn't trust doctors do inject "chemicals" into her children, and showed a (true) story about a child who died of a flu vaccine.
The people opposing vaccination are not evil, they just have not be taught statistics and rational thought. They have no idea how vaccines work. They don't know what sources of information to trust.
Plastics containing just H and C can probably be burned without much processing. The problem is that most recycled plastic is a mix of all sorts, including a lot of popular Cl containing plastics that are really nasty to burn.
If there are sources of sufficiently pure hydrocarbon plastics (polyethylene, polypropylene and the like) and if this is more cost effective and energy effective than other methods, its fine. I expect the bigger problem is the initial purification. Maybe there is a solution to separating out the other plastics?
Sometimes there isn't a clear boarder between sensitive and non-sensitive information. Many people do work at home, or on personal laptops while traveling. While that certainly woudln't include classified information, it might be related to work that is sensitive - sometimes just in work emails.
Often this work is done on people's personal time, so expecting them to go to extra effort to carry additional devices is likely to result in them just not doing the work, and a reduction in productivity.
If I were required to carry a work laptop when on personal travel, I would stop doing work for free when traveling.
Its easy to say "we want everyone to have their personal star-ship by 2030", but if you don't have a plan to get there you just look foolish when you fail.
If there is a general plan, then they should show it. At what rate does solar and wind production need to be ramped up. That tells you about how many factories to build, how many workers etc. Large projects know how to do this.
If it needs new technology then say that: "we've calculated that we can ramp up solar and wind quickly enough but will need *new technology* for energy storage". That tells people what is missing and where to put R&D.
Otherwise, why 10 years? Why not 5, or 1, or tomorrow? What is the argument that 10 years is the right time scale. To me it seems absurdly short - they sent 10 years building a single railway overpass near my house, and 30 rebuilding a single damaged bridge. How can anyone imagine a huge change in US infrastructure in 10 years?
If there is a plan, then lets see it. If not, they are just discrediting legitimate programs to reduce CO2 emissions .
The problem is the crime predicting software may cause police to investigate people against whom there is no solid evidence. Investigation is not harmless -it can cause significant inconvenience and in some cases physical risk. It can also uncover minor crimes that are committed by many people, but usually not noticed. So harm can be caused to people without anything like due process.
Its very difficult to keep bias out of machine learning data sets. If past bias has caused an unusually high number of arrests for some minority group, information that is a proxy for membership in that group will cause more investigations, which will lead to a higher rate of detected crimes and a positive feedback. Its exactly the same sort of feedback that operates with humans, but without the higher reasoning functions that allow some humans to counteract this pattern matching.
Self flying / taxxing aircraft aren't that difficult a problem and they would interact well with automated air traffic control. As long as there are humans though, I think fully automated traffic control would be difficult. Instructions are still given by voice and they can be misunderstood and / or garbled etc.
Automation is also very poor at dealing with unusual situations - say a major earthquake at SFO that might have damaged runways, or a piece of debris falling off of a plane and creating a hazard, or a confused inexperienced (or sometimes experienced) pilot blundering into the wrong airspace. I think some humans will be needed to deal with these sorts of situations.
A company wants verifiable identities on the people who use their site, which will increase the value of the data that company sells to their customers.
Convincing the users (product) to go along is just marketing.
That is true only if you are assuming its the earth that is eclipsing the moon. Other options are however substantially less likely, and less desirable.
If it works as described, and if the cost is reasonable, this would be fantastic for aircraft ice protection. Valuable for small planes that don't have de-ice heat, and for airliners to need less area covered with de-ice gear.
If it works. This type of success has been described before, so I'll wait a bit before I believe it. If it does really work though, its great .
Most big hacks have been due to human factors, not someone breaking zillion-bit encryption.
Computers are very fast in terms of quantitative performance metrics, but I think are less capable of doing useful things that we expected 40 years ago. Probably that is because things like AI turned out to be far harder than most people expected.
Can I have intelligent conversations with my computer: no (no HAL 9000)
Are there regular commercial flights to the moon, and a permanent base on Mars? no. I can't even fly in a supersonic passenger plane anymore.
Fusion reactors? No. Flying cars? No. Undersea cites: No. Humanoid robots helping me round the house: no.
We have fallen far short of what was predicted when I was young. Of course that may be because the predictions were crazy, not because we have in some way failed.
My feeling is that we failed at space. The other things turned out to be harder than we expected.
There are so many attacks on privacy that the public can't resist them all. Its easy to suggest that people not purchase products that spy on them, but when it becomes common for an entire industry (cell phones, TVs, etc), eventually the consumers just give up.
I think we are heading toward 24/7 surveillance that is almost impossible to avoid. That date will then be sold to the highest bidder - or hacked. Companies will use machine learning to look for marketing opportunities. Governments will do the same looking for signs of disloyalty.
Its a rather scary future, but very difficult to avoid.
To get rid of the panopticon, we need to find some other source of revenue. At the moment there is no way to guarantee that if you pay for services you will no longer be tracked, so we are in a sort of latch-up.
People understand this as individuals - they run marathons, climb mountains, learn all sort of new "useless" skills, but we've forgotten that its important to do as a society as well. We are so focused on "efficiency" that we forget to have goals
Groups with obvious CP related names? Maybe, but it could also be an organization that wants to outlaw strong encryption so that they can monitor all communication, and which is safe from prosecution under CP laws.
There are many claims (which of course may be false) that the FBI is the worlds largest distributor of child porn - done of course to attempt to catch (entrap?) peole trading illegal material.
This may legitimately be a bunch of incredibly stupid child pornographers, in which case I hope they spend a good long time in prison. OTHO, its sure convenient for organizations that want to stop encryption.
If these problems apply to payroll and purchasing systems, then its a problem that should be fixed. If they apply to actually missile systems, then of course its a whole different kettle of fish.
I hope and assume any missile systems or classified systems are air-gapped, and things like 2-factor authentication and anti-virus do not apply. Security is guys with guns who shoot anyone who crosses the air-gap without authorization.
I also hope that any report on vulnerabilities of missile systems would be classified.
I'd assume that a machine CBP touched should be immediately destroyed. I have no faith that they haven't installed something or that they may have made security mistakes that allowed the machine to be compromised. I'd probably tell them to keep it.
Yes, its a remarkable technology. You can close the door and avoid distractions. ... or.... I know this is a leap.... open the door and collaborate. You can do *both*.
I've never understood why companies pay workers a ton of money, then put them in situations where they can't work efficiently. Then Apple goes and builds a zillion-dollar headquarters that STILL doesn't provide efficient working spaces.
I'd like to believe that high level managers are just clueless an think that if a bunch of smart engineers talk with each other, flying cars, robots and warp drive spaceships will just appear. I doubt it though - I think its just elitism - an office becomes a mark of status to separate high level managers from the workers.
It would be very difficult to do a valid study. There is a lot of difference between lying about a circle or cross on hidden playing card, and lying about whether you gave the Russians your password, which is again different from whether you cheated on your wife.
In many case truth vs lie may not even be all that well defined.
For those and many other reasons, until I see refereed journal studies, I have 0 confidence in any "lie detector".
At some point (not far away these days) the cooling system starts to become as large as the chips it is cooling, and then there isn't a lot of benefit for higher power densities.
I think the energy per gate switch has to come down. There is a fundamental (I think) limit of several times thermal energy (~1/40ev at room temperature). If you cool below room temperature the Carnot efficiency of the refrigerator keeps you from going much further.
I don't know how close current devices are to KT energy these days.
In a field where personal opinions are as important as they are in HR, its going to be difficult to find an unbiased training set. The resulting AI could easily make strongly biased decisions.
Clearly this is an utterly insane idea if there is any chance of unauthorized parties taking over cars and say driving them into rivers, or each other by the tens of thousands or millions. Talk about a terrorists dream.
So I assume the there is some perfect security solution that makes this all work. Tesla and the like must be much better at security than say the NSA - which got itself badly hacked a while ago.
I'd really like to see all journals be open.
One problem that arises is that science puts a lot of weight on research published in high impact refereed journals. There is a general believe that a paper in Nature or Science or the like has been carefully reviewed. The really is due to the reputation of those journals.
Open journals may eventually gain the same reputation, but it is not an instant process. There are a lot of junk journals out there that will claim to referee, but are really "pay to publish" - anything you send them will be published.
So if science moves away from traditional journals, how can the interested, but non-expert (in the field) public know what has actually been reviewed. How can funding agencies judge the quality of research form publicly funded science?
I expect that in the long term some open journals will develop widely recognized reputations. (remember a reputation just in a field doesn't help with the above problems). I just don't see how to get there quickly.
Is it plausibly useful for very high altitude drones, mars aircraft and the like? What is the effective exhaust velocity? eg is there any regime where it is more efficient than an electric motor and propeller?
Still its a cute concept, even if it isn't practical.