Bit of a quandary. Iâ(TM)m Josef Frisch, a career scientist- but not in climate science. If Iâ(TM)m so confused, how can you know that the climate scientists arenâ(TM)t also confused. Why do you believe what you believe?
Also I think that if we want people to respect / believe science, then science needs to be separate from politics. It needs to not take sides, but to provide politically unbiased information for people to use to set policy.
One of the largest objections to current climate science is the claim that it has a liberal bias. We need to be very careful not to let that be true. Unfortunately science isn't immune to political pressure - I see it even in the non-controversial things that I work on. It takes extreme vigilance to protect science from political pressure (from either side) on the politically sensitive issues.
So you think we know everything we need to know about climate science and don't need to study it further? If not, then its not "settled".
Otherwise you have to clarify what you think is settled? That humans cause climate change - of course - but the question is how much and the range of predictions from the IPCC is large.
I think we need to continue to do climate science to provide better input for policy decisions.
BTW - at the end of the last ice age the Arctic *did* warm all by itself - or really in response to the variations in the earths tilt and orbit. Climate does change without human action. The question is the extent to which human action effect things. The answer looks like human effects have been quite large relative to other changes on the scale of the last century or two, but it is extremely complex to model.
Science can tell us what will happen in different scenarios, but it can't tell us what we SHOULD do about it because science says nothing about what our goals are. There world will be different under different policies, but its not the job of science to tell us which of those futures are better.
Should we kill or sterilize everyone with genetic defects? Science will say that course of action will very gradually decrease the number of defects in the population, and the Nazis used this as an argument that we SHOULD take that action. Most people today, myself included, think that is a terrible course of action because we are not trying to optimize the genetics of the human race, that isn't the goal.
In the case of climate, I think that science provides input that makes a political decision to reduce CO2 emissions a good choice, but it remains a political, not scientific choice.
Science isn't a yes / no. Certainly we don't completely understand climate.
If you want a yes / no, you need to ask a very specific question:
Does human activity affect climate? Yes - obviously.
Should we we reduce CO2 emissions? That isn't a "science" question, it is a political question that takes (or should take) as inputs climate models and economic models.
The real questions are things like:
For various CO2 emission scenarios, what are the likely ranges of sea level, and climate changes in different parts of the world. These are being worked on, but there is still a large range in the simulation predictions.
Are we missing any important inputs to climate? (like the cosmic ray / solar wind effect on cloud seeding issue).
If the science were settled there would be no point spending more effort on it. (Newtonian mechanics is "settled", no one does research on Newtonian mechanics).
I agree if NOAA is funded to do the work that NASA was doing. If NASA's climate effort is cut, but NOAA's isn't increased, I think that is a mistake.
I not one of the "the science is settled" people - its a complex problem and one with potentially very large impact, so we need to put a lot of effort into studying it.
What I meant was that if you have sufficient family ties, you can be wealthy, and someone will give you a "job" if you want it.
Luck - well a lottery solves a lot. (or a virtual lottery where you join a company and even though you don't really do much, you end up with very valuable stock).
I think that peopleâ(TM)s impact on productivity varies a lot more than 10x.
Once I come is determined by something others than the free market you get all the standard command economy problems with who decides what people are paid.
There are technical solutions, but unless they are automatic, or at least extremely simple, the general public will not use them. I think that a right to privacy should not be restricted to people with substantial computer skills.
I see a number of dangers form the compilation of personal data from large numbers of people and I would want that to stop for everyone even if I were savvy enough to protect myself.
I completely agree that encryption is a big part of the solution, I just think that it needs to be mandated to be applied to everyone.
What I mean by "root source of trust" is something people agree upon to be true. For instance if there were a trusted organization to verify that this DNS service is what it claims to be (discounting legitimate bugs), and not a front for some organization intent on using it to collect information
It may just be my background but science feels a little different: different scientific ideas interlock with each other. It would be difficult to fake a significant branch of science because all the interfaces with other types of science would be off. That seems different from trusting a hosting site, or an implementation of an encryption or communication system which could itself be flawed without other major consequences.
I'm a scientist, so my viewpoint may be very biased on this. If I were a computer scientist, I might have a very different idea of what could and couldn't be trusted.
Actually I've sometimes wondered if that was one of the functions of the medieval church: they served as an agreed-upon source of truth. That agreement was useful to society even if they were not actually truthful .
We have the technology top ensure that everyone spends their time doing productive work. The only problem is defining "productive" . By what measure is traveling in a flying car more productive than watching a porn video? What is the end goal that we are trying to optimize by banning porn?
Depends. Putting sensitive information on someone else's servers is risky, but there is a lot of information that is NOT sensitive. I use dropbox extensively as a convenient way to access files from multiple locations and to share files. I only put files there that I woudln't mind becoming completely public. I'm really not worried if someone steals the slides for a talk I've already given in public, or gets my vacation photos.
My secret plans for world domination don't go on dropbox.
Good question. Are the algorithms in, for example, aircraft navigation systems public, or are they only required to demonstrate the required performance?
It would of course be interesting to know what suite of sensors they are using - but that is presumably proprietary. The courts could request the evidence.
Probably though, until there are specific safety standards for sensors in AVs, the standard to use now is to compare with a human driver. From the video, I think the majority of human drivers would have failed to avoid that collision.
Completely agree. Humans are just not good at concentrating attention all of the time with nothing to do. Combine that with the low visible contrast of the pedestrian, and I don't think its reasonable to think that a backup drive could have been attentive enough to take over.
There is still value in having backup drives to deal with situations that have somewhat longer timescales.
1) Why is quasar alignment surprising when you have large scale inhomogeneties that eventually form large scale clusters? Gravity enhances statistical fluctuations in the original big bang, - this results in galactic clusters.
2) "apparent" interactions needs to be shown with real statistics. Is there a paper- including gravitational lensing?
3) Need a real statistical analysis. Not "numerous".
4) Statistics? Quasars are known at large distances due to gravatational lensing from closer galaxies. What other explanation is there?
5) Seemingly? How can you tell it isn't just random overlap? Again, lensing of quasars seems to rule the nearby quasar theory out completely.
6) Yes, ultra-relativistic jets can appear to exhibit superluminal motion. Well understood since at least the 80s. Its a well understood and modeled optical illusion.
7) The universe is obviously not uniform (rocks vs empty space). Big bang predicts statistical fluctuations at different wavelengths. Is this outside of that prediction?
8) What does time dilation of quasar variation mean? How do you know the original timescale? Fluctuations have been seen delayed by lensing
9) paper? That is an interesting one.
10) Again, big bang doesn't say the universe is uniform. It allows statistical fluctuations. (at least big bang theory in the last 30 years. Scientists have been studying the flutuations for decades now, and seen clear signs of inflation in the statistical variations. This feels like someone looking at decades old data and theories.
11) don't know who these people are.
Overall: if quasars are local, how can gravitational lensed images be explained. Many years ago people thought quasars might be local, but a huge set of experiments since then has show that that is not the case
We know a LOT more about the universe than we did 50 years ago. (That darn well better be the case for the billions we have spent on basic science). When we have the right answers, breakthroughs are going to be uncommon.
When I was a grad student, astrophysics data was pretty thin and there were lots of discoveries to be made: inflation, dark matter, dark energy, quasar engines, etc etc. Now we have done all the "easy" observations that anyone can think of, so new data is very expensive. If anyone can think of a new way to make astrophysical measurements, I'm all ears. (or eyes, or whatever appendage is appropriate).
I think the place to look for breakthroughs is in new science, not fields that have been extensively studied like astrophysics.People are doing laboratory scale experiments in quantum entanglement for example, so some breakthrough there is quite possible. Gene sequencing is providing a new set of data for evolution science, paleontology and human origins. Improvements in computation are benefiting a variety of fields. Young scientists are strongly encouraged to get into these fields.
What is not going to work is for people to try to make "breakthroughs" in fields where vast amounts of effort have already been spend, unless they are willing to put in the time to understand what is already know. Over lunch I was laughing with a colleague about how completely absurd the idea was that the universe isn't expanding - it is supported by such a huge wealth of data. At the very least, how could someone explain the supernova red-shift data without cosmological expansion?
I do agree with you that science has become very risk adverse. The problem is that the experiments have become so expensive that governments are worried about looking bad if they fail. Imagine he news media having a field day with a government project that spent a billion dollars and failed to measure anything new.
Which paper has statistics? (can you post the link). The original one doesn't have any statistical analysis that I can find. There are lots of objects so there will be lots of cases where distant objects seen to be in line with near ones.
Slashdot Mirror
NOAA could use commercial launches-if they were funded to do so
I donâ(TM)t think I disagreed with that. I only disagree that the science is done (settled).
Bit of a quandary. Iâ(TM)m Josef Frisch, a career scientist- but not in climate science. If Iâ(TM)m so confused, how can you know that the climate scientists arenâ(TM)t also confused. Why do you believe what you believe?
Agreed
Also I think that if we want people to respect / believe science, then science needs to be separate from politics. It needs to not take sides, but to provide politically unbiased information for people to use to set policy.
One of the largest objections to current climate science is the claim that it has a liberal bias. We need to be very careful not to let that be true. Unfortunately science isn't immune to political pressure - I see it even in the non-controversial things that I work on. It takes extreme vigilance to protect science from political pressure (from either side) on the politically sensitive issues.
So you think we know everything we need to know about climate science and don't need to study it further? If not, then its not "settled".
Otherwise you have to clarify what you think is settled? That humans cause climate change - of course - but the question is how much and the range of predictions from the IPCC is large.
I think we need to continue to do climate science to provide better input for policy decisions.
BTW - at the end of the last ice age the Arctic *did* warm all by itself - or really in response to the variations in the earths tilt and orbit. Climate does change without human action. The question is the extent to which human action effect things. The answer looks like human effects have been quite large relative to other changes on the scale of the last century or two, but it is extremely complex to model.
How is a policy decision a science question?
Science can tell us what will happen in different scenarios, but it can't tell us what we SHOULD do about it because science says nothing about what our goals are. There world will be different under different policies, but its not the job of science to tell us which of those futures are better.
Should we kill or sterilize everyone with genetic defects? Science will say that course of action will very gradually decrease the number of defects in the population, and the Nazis used this as an argument that we SHOULD take that action. Most people today, myself included, think that is a terrible course of action because we are not trying to optimize the genetics of the human race, that isn't the goal.
In the case of climate, I think that science provides input that makes a political decision to reduce CO2 emissions a good choice, but it remains a political, not scientific choice.
Science isn't a yes / no. Certainly we don't completely understand climate.
If you want a yes / no, you need to ask a very specific question:
Does human activity affect climate? Yes - obviously.
Should we we reduce CO2 emissions? That isn't a "science" question, it is a political question that takes (or should take) as inputs climate models and economic models.
The real questions are things like:
For various CO2 emission scenarios, what are the likely ranges of sea level, and climate changes in different parts of the world. These are being worked on, but there is still a large range in the simulation predictions.
Are we missing any important inputs to climate? (like the cosmic ray / solar wind effect on cloud seeding issue).
If the science were settled there would be no point spending more effort on it. (Newtonian mechanics is "settled", no one does research on Newtonian mechanics).
I agree if NOAA is funded to do the work that NASA was doing. If NASA's climate effort is cut, but NOAA's isn't increased, I think that is a mistake.
I not one of the "the science is settled" people - its a complex problem and one with potentially very large impact, so we need to put a lot of effort into studying it.
What I meant was that if you have sufficient family ties, you can be wealthy, and someone will give you a "job" if you want it.
Luck - well a lottery solves a lot. (or a virtual lottery where you join a company and even though you don't really do much, you end up with very valuable stock).
Usually though you need some of all of the above.
Talent, hard work, luck, family all are important for success.
Luck or family by themselves will do it. The others need some combination to work.
I think that peopleâ(TM)s impact on productivity varies a lot more than 10x.
Once I come is determined by something others than the free market you get all the standard command economy problems with who decides what people are paid.
There are technical solutions, but unless they are automatic, or at least extremely simple, the general public will not use them. I think that a right to privacy should not be restricted to people with substantial computer skills.
I see a number of dangers form the compilation of personal data from large numbers of people and I would want that to stop for everyone even if I were savvy enough to protect myself.
I completely agree that encryption is a big part of the solution, I just think that it needs to be mandated to be applied to everyone.
What I mean by "root source of trust" is something people agree upon to be true. For instance if there were a trusted organization to verify that this DNS service is what it claims to be (discounting legitimate bugs), and not a front for some organization intent on using it to collect information
It may just be my background but science feels a little different: different scientific ideas interlock with each other. It would be difficult to fake a significant branch of science because all the interfaces with other types of science would be off. That seems different from trusting a hosting site, or an implementation of an encryption or communication system which could itself be flawed without other major consequences.
I'm a scientist, so my viewpoint may be very biased on this. If I were a computer scientist, I might have a very different idea of what could and couldn't be trusted.
Actually I've sometimes wondered if that was one of the functions of the medieval church: they served as an agreed-upon source of truth. That agreement was useful to society even if they were not actually truthful .
I wish we could trust the government. I'd hope that in a democracy we could and they would provide that - but sadly we can't.
With this and all other attempts to provide privacy or security, what chain of trust allows me to believe that this is actually private or secure.
Surely there are many organizations with the resources to flood Slashdot with posts assuring me that this, or any other service, is secure.
Is TOR secure, or a NSA honeypot? How could I possibly know? Without personally having deep technical expertise, how can I trust anything.
An comments about tinfoil hats could be legit, or yet more planted posts.
We need a root source of trust or everything else falls apart.
We have the technology top ensure that everyone spends their time doing productive work. The only problem is defining "productive" . By what measure is traveling in a flying car more productive than watching a porn video? What is the end goal that we are trying to optimize by banning porn?
Depends. Putting sensitive information on someone else's servers is risky, but there is a lot of information that is NOT sensitive. I use dropbox extensively as a convenient way to access files from multiple locations and to share files. I only put files there that I woudln't mind becoming completely public. I'm really not worried if someone steals the slides for a talk I've already given in public, or gets my vacation photos.
My secret plans for world domination don't go on dropbox.
Good question. Are the algorithms in, for example, aircraft navigation systems public, or are they only required to demonstrate the required performance?
It would of course be interesting to know what suite of sensors they are using - but that is presumably proprietary. The courts could request the evidence.
Probably though, until there are specific safety standards for sensors in AVs, the standard to use now is to compare with a human driver. From the video, I think the majority of human drivers would have failed to avoid that collision.
Completely agree. Humans are just not good at concentrating attention all of the time with nothing to do. Combine that with the low visible contrast of the pedestrian, and I don't think its reasonable to think that a backup drive could have been attentive enough to take over.
There is still value in having backup drives to deal with situations that have somewhat longer timescales.
1) Why is quasar alignment surprising when you have large scale inhomogeneties that eventually form large scale clusters? Gravity enhances statistical fluctuations in the original big bang, - this results in galactic clusters.
2) "apparent" interactions needs to be shown with real statistics. Is there a paper- including gravitational lensing?
3) Need a real statistical analysis. Not "numerous".
4) Statistics? Quasars are known at large distances due to gravatational lensing from closer galaxies. What other explanation is there?
5) Seemingly? How can you tell it isn't just random overlap? Again, lensing of quasars seems to rule the nearby quasar theory out completely.
6) Yes, ultra-relativistic jets can appear to exhibit superluminal motion. Well understood since at least the 80s. Its a well understood and modeled optical illusion.
7) The universe is obviously not uniform (rocks vs empty space). Big bang predicts statistical fluctuations at different wavelengths. Is this outside of that prediction?
8) What does time dilation of quasar variation mean? How do you know the original timescale? Fluctuations have been seen delayed by lensing
9) paper? That is an interesting one.
10) Again, big bang doesn't say the universe is uniform. It allows statistical fluctuations. (at least big bang theory in the last 30 years. Scientists have been studying the flutuations for decades now, and seen clear signs of inflation in the statistical variations. This feels like someone looking at decades old data and theories.
11) don't know who these people are.
Overall: if quasars are local, how can gravitational lensed images be explained. Many years ago people thought quasars might be local, but a huge set of experiments since then has show that that is not the case
11)
We know a LOT more about the universe than we did 50 years ago. (That darn well better be the case for the billions we have spent on basic science). When we have the right answers, breakthroughs are going to be uncommon.
When I was a grad student, astrophysics data was pretty thin and there were lots of discoveries to be made: inflation, dark matter, dark energy, quasar engines, etc etc. Now we have done all the "easy" observations that anyone can think of, so new data is very expensive. If anyone can think of a new way to make astrophysical measurements, I'm all ears. (or eyes, or whatever appendage is appropriate).
I think the place to look for breakthroughs is in new science, not fields that have been extensively studied like astrophysics.People are doing laboratory scale experiments in quantum entanglement for example, so some breakthrough there is quite possible. Gene sequencing is providing a new set of data for evolution science, paleontology and human origins. Improvements in computation are benefiting a variety of fields. Young scientists are strongly encouraged to get into these fields.
What is not going to work is for people to try to make "breakthroughs" in fields where vast amounts of effort have already been spend, unless they are willing to put in the time to understand what is already know. Over lunch I was laughing with a colleague about how completely absurd the idea was that the universe isn't expanding - it is supported by such a huge wealth of data. At the very least, how could someone explain the supernova red-shift data without cosmological expansion?
I do agree with you that science has become very risk adverse. The problem is that the experiments have become so expensive that governments are worried about looking bad if they fail. Imagine he news media having a field day with a government project that spent a billion dollars and failed to measure anything new.
Which paper has statistics? (can you post the link). The original one doesn't have any statistical analysis that I can find. There are lots of objects so there will be lots of cases where distant objects seen to be in line with near ones.
How can this explain the correlation of supernova brightness with redshift https://ned.ipac.caltech.edu/l...
or gravitational lensing of distant quasars by nearby galaxies.
Or gravitational lensing of the CMB background?
Its a fun discussion why stop now?