I didn't realize having $1 million dollars of insurance counted as a skill. But certainly if we forced burger flippers to buy the equivalent insurance, they would also possess this skill.
That appears to be an example of a person who created fake registrations in order to keep his job at acorn, not an example of thousands of fraudulent votes actually being cast.
Do you have any actual data showing the amount of actual voter fraud (i.e. fraudulent votes being cast?
Furthermore, is there any evidence that shows democrats are doing more voter fraud than republicans? I remember a story a while back about republicans sending fake pamphlets to democrats reminding them to vote on the wrong say, as well as pamphlets warning Mexican Americans not to vote or or risk deportation.
And these are the sorts of voter fraud that would not even be helped by voter ID laws.
Do you have any evidence for the "enourmous numbers of bogus voter registrations"?
Is there any data you can cite?
And a philosophical question: If the situation were reversed, and there were many more republicans without ID, do you think the republicans would still be pushing hard for voter ID laws?
There is more to software than algorithms. Algorithms are a tool (e.g. like a printer, etc). The algorithm takes input and generates an output. You can make a printer print out a KKK written word doc, but that doesn't make the printer racist. When I say "I don't know how to make a racist algorithm", it's like saying "I don't know how to make a racist printer". I am open to suggestions.
I am not saying it's impossible that a racist printer could exist. What I am saying is that they must be pretty fucking rare if I (as a printer expert), can't even imagine how it could be done in practice.
And in this respect, I think algorithms are very similar to math equations. I can't imagine how to make a racist math equation. That doesn't mean it can't be done.
This is in contrast to making a racist child who then becomes a racist adult. I don't claim to have a 100% effective way to do it, but I've seen examples of lot's of people succeeding at it. I know what a racist person is. I don't even know what a racist equation, or algorithm, or a racist printer even is.
A man falls into a volcano. Who is at fault? The man for not staying a safe distance away from the volcano, or the owners of the property for not properly safeguarding a dangerous volcano.
I don't know who is at fault. It's for the courts to decide using relevant laws and contracts, etc. But this isn't a new problem.
Determining fault is pretty much all courts really do.
I do know what you were getting at. But you apparently did not know what I was getting at, which is that just because something is created by humans does not imply that it can/will be imbued with the racism of that human.
Math equations are a good example of this non-implication. I'm sure lots of mathematicians have been racist and yet there are no racist math equations (at least none that I have seen).
I am a software developer and I'm not even really sure what a racist computer algorithm would look like. If you can give me an example of one, go right ahead.
There was a time when sending an email was vastly more complicated than sending a physical letter via the post office. To those with no imagination, it was obvious that sending a letter through the post office was inherently easier than any sort of computer mailing system ever could be, because sending a letter is so simple, and computers are so complicated.
If you're worried about people not knowing there's an election coming up, and not bothering to get an ID (really? you can't go to the doctor, fill a prescription, collect a welfare check, or much of ANYTHING else with already having an ID),
I'm not worried because I'm not a democrat, and don't have any interest in helping democrats win elections.
What I am saying is that voter fraud is currently not a problem. There just aren't significant number of fraudulent votes. It is pretty clear that the problem is inflated by republican politicians and strategists to get voter ID laws to help them win elections.
then why not encourage the Democrats to apply the same level of effort they put into the shady practices described above, and focus it instead on getting that rare person who never sees a doctor, never gets a prescription, collects no government benefits of any kind, doesn't work (but whom you seem to suggest none the less are a large voting block) and, with YEARS to work with between elections... just getting them an ID?
I'm not saying it's hard to get an ID. I'm saying that t will result in less votes for democrats, and everyone knows that. The republicans know it and the democrats know that, and that's all both parties care about.
If the situation were flipped, and there was a law to change voting registration to be done on instagram or twitter or something, then you'd have republicans fighting it and democrats pushing it. Neither party gives a damn about fairness.
I don't disagree with the idea of preventing voter fraud nor government issued ID's as the tool to accomplish that task. That said it is pretty obvious that the main proponents of voter laws are Republicans because they know it will benefit them in elections, and the main opponents of voter laws are democrats because they know it will not benefit them in elections. Neither side cares about fairness, they only care about winning.
If voter fraud were a big problem, I think the disparity in outcome would not trump the need for legitimacy. But the gap in number of democrats without photo ID's vs republicans without photo ID's suggests that democrats will lose way more votes from voter ID laws than anything close the the amount of voter fraud going on.
It's not that democrats can't get free state ID's, it's that we can be sure that some % of everyone without IDs will fail to get them before the election, and the democrats have a much larger pool of people that can potentially fall into this category.
Some things are probably just harder to classify correctly than others. There are probably lots of examples of different animals and colors that are harder or easier to classify.
I'm sure if you you stood against a background that was the same color as your skin, the algorithm would do a better job of detecting your partner
You can probably run those pictures through an edge detection algorithm and see that depending on the background, people of different skin tones stand out more. This doesn't mean the algorithm is racist. This is just a property of the images.
If I painted my face orange, I'll bet there is a higher chance that I will be detected as an orangutan.
Yes it's important to have good training data, but that doesn't change the fact that some differences require more training to differentiate.
Some asshole must have changed wikipedia to make you wrong. It says:
The Hominidae (/hmndi/), also known as great apes,[notes 1] or hominids, form a taxonomic family of primates, including four extant genera: orangutans (Pongo) with two species extant; gorillas (Gorilla) with two species; chimpanzees (Pan) with two species; and humans (Homo) with one species.[1]
You'd better go in there and correct it to say say humans are not apes.
Their algorithm incorrectly identified one member of the Homininae family (gorillas) with another member (humans) sharing at least 95% of the same DNA. Yeah they are fucking idiots. How well does your algorithm work?
I think inspiring people to go into STEM is a very noble and worthwhile goal. I just have a hard time believing that manned missions are necessarily better for this. I don't dispute that manned missions in the past have attracted more attention than unmanned missions, but they were also missions that spent much more money, and had the largest engines ever created by man, etc, etc.
There were more than enough things that needed to be engineered during the spaceflights during the 60's without removing humans from the space ship. This necessity for humans in the spaceship certainly added to the suspense, but without that necessity, I feel like manned missions will just seem like a foolish rather than necessary (and it will be true).
Do we still need to beat the Russians in space? Not really, but even if we did in 2015, I don't think knowingly risking American lives unnecessarily was or ever will be popular.
I think also there was probably a lot of value in seeing a human being on the moon, over simply watching a video feed from a machine, just in terms of believability. I think we (most of us anyway), are passed the point of needing to see a guy in a space suit on mars to believe that the video feed isn't fake.
Besides, even if a bunch of astronauts go to mars, it's not like I am on Mars. I am still just seeing pictures from earth. Those pictures just won't have any people in them (initially).
I think maybe adjusting people's expectations for what science "looks like" is a good thing. You don't need to see atoms with your own eyes for them to be real. It's ok if we are only seeing a projection of reality that a machine is generating for human eyes. That doesn't diminish the accomplishment.
1. A human can only do what it's programming allows as well, even if that includes noticing novel phenomena.
2. A human being is pretty good at multitasking. We almost always use machines when we need something that can keep it's focus on the task at hand without being distracted by other things. You can make a machine that can multitask as well, but often it makes more sense to just make 2 machines. We can have one machine doing the experiment with 100% concentration, and one just looking for novel phenomena with 100% concentration. In fact it might be better to have a machine that can't be distracted from it's job of looking for things to be distracted by.
Some discoveries. Very slowly. With very limited flexibility and substantially reduced spinoff benefits. Robots are a great way to explore some things but they are not the preferred way to explore everything.
Preferred by who? The scientists? The tax payers? The explorers?
How digestible the plant is by people in space to start with. How the plant interacts with humans in a different environment for another. You cannot discover a lot of things that relate to humans without a human being present.
When it comes time to actually eat a plant grown on mars, then sure we will need a person to do that, but even then it will first be analyzed by machines to make sure it has only safe chemical compounds.
Who said anything about making something deficient? Strawman argument you have there my friend. Don't try to put words in my mouth.
A robot that can't do something is deficient of that ability. By having a mission where humans rather than machines have the responsibility to accomplish certain tasks, you are intentionally making those machines deficient of those abilities. I am saying that as an engineer (someone interested in engineering), the possibility of making something that can do it's job with less human intervention is more exciting than if it requires humans next to it to make it work right. I am disputing your claim that manned missions generate more interest in science and engineering. I think they generate more interest in being an astronaut.
Of course we'll invent good spinoff technologies. But they will be DIFFERENT technologies. There are some technologies that will only and can only be developed if you plan to send people.
And you know exactly which spinoff technologies will be invented, and therefore know that the spinoff technologies of manned mars missions will be more useful than the spinoff technologies of unmanned mars missions?
And this robot vs human thing is a false dilemma. It does not have to be an either/or proposition. We can and should do both.
We were never doing *only* manned missions. There were always machines doing most of the work in space. The difference is that now *all* the space work can be done by machines. Or another way to look at it is that now *all* the human work can be done on the ground.
There already is a tangible benefit to sending people. I outlined a large number of them. It will also take us decades to get to the point where sending people is a realistic option.
The tangible benefits you listed were all related to learning about how to send people to mars. That's like saying the tangible benefit to sky diving is learning how to sky dive. Fine maybe learnign how to sky dive is a benefit. What I am saying is that we should learn to sky dive when it benefits us to be able to sky dive (apart form the benefit of simply knowing how to sky dive).
If all we cared about were the benefits of spinoff technologies. We could simply invest that money in those technologies and probably do it more efficiently than through this indirect approach.
I never claimed that we should. Not sure who's argument you are responding to but it certainly isn't mine.
I never claimed this was your argument. I think you have straw man sensitivity. This was part of my argument to illustrate when *I* think we should send humans and when I think we shouldn't.
How much more money did NASA spend on the Apollo missions than the mars rover missions? I find it ironic that in your example of a human doing more than a machine they are still doing it in a machine.
Sending a big rover that can drive longer ranges is just a matter of spending more money on bigger rockets and more fuel to deliver more cargo into space.
We couldn't go anywhere in space without the precision of computer control systems.
What is this team of scientists going to do on mars? Look at things with their eyeballs and touch things with their fingers? No they are going to be using instruments to collect data (whether it's numbers, images, etc). All that stuff can be sent back to earth for analysis.
It's like saying "We *need* human pilots because a little quadcopter I bought at toys'r'us doesn't have the range of a Boeing 787". Even though not only can an autopilot be installed in a 787, but they already actually have far more sophisticated autopilots than some toy quadcopter.
For the price of having a team of scientists on mars, we could probably have hundreds of rovers and teams on earth supporting them. Or we could probably just make far better rovers.
What makes a human better than a rover? That a human can walk faster than a rover can roll? That he/she can climb over terrain better? Those are all things that rovers will can/will get better at (if we are willing to spend the money), where humans will never really improve.
Rovers are probably not going to be good at making high level decisions (e.g. how should I conduct this experiment?), but those sorts of decisions don't need low latency. The decisions that do need low latency (low level decisions (e.g. how should I avoid this rock), are already getting to the point to where the are close to being better than humans (especially in an environment that humans aren't used to).
There really is no reason that high level decisions need to be made on mars. And high level decision making is really the only thing that humans will do better than robots for the foreseeable future.
You get the most sophisticated tool we possess on Mars. One that can make discoveries orders of magnitude faster than any other tool we possess.
The most sophisticated tool we possess, discovered a way to make discoveries without leaving earth.
What can a human being learn about botany in space that a human can't learn on earth by controlling a robot botanist?
and countless other subjects not relevant to mechanical robots
Nothing is relevant to robots. They are robots. Everything they do is something that is relevant to humans. Artificial sensors are better at detecting things than human beings (even the things that are only relevant to humans). A robot will know the CO2 level in a room better than a human ever will. That's we we use instruments to measure CO2 levels and don't just ask people how much it feels like there is.
You'll also inspire a lot of people to get into science and engineering - far more than any robot mission ever could.
As an engineer, (and not an astronaut), I think I am far more interested in making the thing that actually goes to mars and does the work, rather than making something that is so deficient that it requires a human being to be in close physical proximity to operate it.
I think we will invent good spinoff technology regardless of whether we send humans or robots. In fact I would say the *best* spinoff technology to come from the space race were the advances in automation.
You know there used to be a time when we planned (and the russians actually did) send manned spy satellites into space. The job of the person on board was to point the spy camera at interesting things to spy on, and also use the on-board weapon systems to shoot downl other spy satellites. Before we actually finished ours, someone (very smart) realized that the future was to send unmanned spy satellites. "How will the machine possibly do as good of a job as a human?" people said. It turns out that those people just failed to understand what was possible through automation.
I'm not saying we shouldn't have people on mars. We should when it benefits us. Automation removes the *need* to put humans on mars to actually do things on mars. We should go when there is a tangible benefit to going.
We shouldn't send people to mars to repair robots. Robots can repair robots. We shouldn't send people on mars to operate instruments. Robots can operate instruments. We shouldn't send people to mars to point cameras. We shouldn't send people to mars to lift heavy shit. We shouldn't send people to mars to push buttons on a computer. These are all reasons we used to send peopel to places they didn't want (but needed) to go.
We should send people to mars when it is better for those people to be on mars than on earth.
Slashdot Mirror
They also tried to prevent open source encryption software from being posted to the internet using ITAR as well. Look how well that turned out.
Sure
I didn't realize having $1 million dollars of insurance counted as a skill. But certainly if we forced burger flippers to buy the equivalent insurance, they would also possess this skill.
That appears to be an example of a person who created fake registrations in order to keep his job at acorn, not an example of thousands of fraudulent votes actually being cast.
Do you have any actual data showing the amount of actual voter fraud (i.e. fraudulent votes being cast?
Furthermore, is there any evidence that shows democrats are doing more voter fraud than republicans? I remember a story a while back about republicans sending fake pamphlets to democrats reminding them to vote on the wrong say, as well as pamphlets warning Mexican Americans not to vote or or risk deportation.
And these are the sorts of voter fraud that would not even be helped by voter ID laws.
Do you have any evidence for the "enourmous numbers of bogus voter registrations"?
Is there any data you can cite?
And a philosophical question: If the situation were reversed, and there were many more republicans without ID, do you think the republicans would still be pushing hard for voter ID laws?
There is more to software than algorithms. Algorithms are a tool (e.g. like a printer, etc). The algorithm takes input and generates an output. You can make a printer print out a KKK written word doc, but that doesn't make the printer racist. When I say "I don't know how to make a racist algorithm", it's like saying "I don't know how to make a racist printer". I am open to suggestions.
I am not saying it's impossible that a racist printer could exist. What I am saying is that they must be pretty fucking rare if I (as a printer expert), can't even imagine how it could be done in practice.
And in this respect, I think algorithms are very similar to math equations. I can't imagine how to make a racist math equation. That doesn't mean it can't be done.
This is in contrast to making a racist child who then becomes a racist adult. I don't claim to have a 100% effective way to do it, but I've seen examples of lot's of people succeeding at it. I know what a racist person is. I don't even know what a racist equation, or algorithm, or a racist printer even is.
A man falls into a volcano. Who is at fault? The man for not staying a safe distance away from the volcano, or the owners of the property for not properly safeguarding a dangerous volcano.
I don't know who is at fault. It's for the courts to decide using relevant laws and contracts, etc. But this isn't a new problem.
Determining fault is pretty much all courts really do.
I do know what you were getting at. But you apparently did not know what I was getting at, which is that just because something is created by humans does not imply that it can/will be imbued with the racism of that human.
Math equations are a good example of this non-implication. I'm sure lots of mathematicians have been racist and yet there are no racist math equations (at least none that I have seen).
I am a software developer and I'm not even really sure what a racist computer algorithm would look like. If you can give me an example of one, go right ahead.
There was a time when sending an email was vastly more complicated than sending a physical letter via the post office. To those with no imagination, it was obvious that sending a letter through the post office was inherently easier than any sort of computer mailing system ever could be, because sending a letter is so simple, and computers are so complicated.
Algorithms are comprised only of math?
Did I say that?
If you're worried about people not knowing there's an election coming up, and not bothering to get an ID (really? you can't go to the doctor, fill a prescription, collect a welfare check, or much of ANYTHING else with already having an ID),
I'm not worried because I'm not a democrat, and don't have any interest in helping democrats win elections.
What I am saying is that voter fraud is currently not a problem. There just aren't significant number of fraudulent votes. It is pretty clear that the problem is inflated by republican politicians and strategists to get voter ID laws to help them win elections.
then why not encourage the Democrats to apply the same level of effort they put into the shady practices described above, and focus it instead on getting that rare person who never sees a doctor, never gets a prescription, collects no government benefits of any kind, doesn't work (but whom you seem to suggest none the less are a large voting block) and, with YEARS to work with between elections ... just getting them an ID?
I'm not saying it's hard to get an ID. I'm saying that t will result in less votes for democrats, and everyone knows that. The republicans know it and the democrats know that, and that's all both parties care about.
If the situation were flipped, and there was a law to change voting registration to be done on instagram or twitter or something, then you'd have republicans fighting it and democrats pushing it. Neither party gives a damn about fairness.
Math equations are created by humans too. Have you ever seen a racist math equation?
I don't think you can determine if someone is American just by looking at them. That's nationalist.
I don't disagree with the idea of preventing voter fraud nor government issued ID's as the tool to accomplish that task. That said it is pretty obvious that the main proponents of voter laws are Republicans because they know it will benefit them in elections, and the main opponents of voter laws are democrats because they know it will not benefit them in elections. Neither side cares about fairness, they only care about winning.
If voter fraud were a big problem, I think the disparity in outcome would not trump the need for legitimacy. But the gap in number of democrats without photo ID's vs republicans without photo ID's suggests that democrats will lose way more votes from voter ID laws than anything close the the amount of voter fraud going on.
It's not that democrats can't get free state ID's, it's that we can be sure that some % of everyone without IDs will fail to get them before the election, and the democrats have a much larger pool of people that can potentially fall into this category.
Some things are probably just harder to classify correctly than others. There are probably lots of examples of different animals and colors that are harder or easier to classify.
I'm sure if you you stood against a background that was the same color as your skin, the algorithm would do a better job of detecting your partner
You can probably run those pictures through an edge detection algorithm and see that depending on the background, people of different skin tones stand out more. This doesn't mean the algorithm is racist. This is just a property of the images.
If I painted my face orange, I'll bet there is a higher chance that I will be detected as an orangutan.
Yes it's important to have good training data, but that doesn't change the fact that some differences require more training to differentiate.
We should throw in some white colored apes that google can mistake for white people, and then it will be fair, and everyone can stop being outraged.
Some asshole must have changed wikipedia to make you wrong. It says:
The Hominidae (/hmndi/), also known as great apes,[notes 1] or hominids, form a taxonomic family of primates, including four extant genera: orangutans (Pongo) with two species extant; gorillas (Gorilla) with two species; chimpanzees (Pan) with two species; and humans (Homo) with one species.[1]
You'd better go in there and correct it to say say humans are not apes.
Their algorithm incorrectly identified one member of the Homininae family (gorillas) with another member (humans) sharing at least 95% of the same DNA. Yeah they are fucking idiots. How well does your algorithm work?
I think inspiring people to go into STEM is a very noble and worthwhile goal. I just have a hard time believing that manned missions are necessarily better for this. I don't dispute that manned missions in the past have attracted more attention than unmanned missions, but they were also missions that spent much more money, and had the largest engines ever created by man, etc, etc.
There were more than enough things that needed to be engineered during the spaceflights during the 60's without removing humans from the space ship. This necessity for humans in the spaceship certainly added to the suspense, but without that necessity, I feel like manned missions will just seem like a foolish rather than necessary (and it will be true).
Do we still need to beat the Russians in space? Not really, but even if we did in 2015, I don't think knowingly risking American lives unnecessarily was or ever will be popular.
I think also there was probably a lot of value in seeing a human being on the moon, over simply watching a video feed from a machine, just in terms of believability. I think we (most of us anyway), are passed the point of needing to see a guy in a space suit on mars to believe that the video feed isn't fake.
Besides, even if a bunch of astronauts go to mars, it's not like I am on Mars. I am still just seeing pictures from earth. Those pictures just won't have any people in them (initially).
I think maybe adjusting people's expectations for what science "looks like" is a good thing. You don't need to see atoms with your own eyes for them to be real. It's ok if we are only seeing a projection of reality that a machine is generating for human eyes. That doesn't diminish the accomplishment.
1. A human can only do what it's programming allows as well, even if that includes noticing novel phenomena.
2. A human being is pretty good at multitasking. We almost always use machines when we need something that can keep it's focus on the task at hand without being distracted by other things. You can make a machine that can multitask as well, but often it makes more sense to just make 2 machines. We can have one machine doing the experiment with 100% concentration, and one just looking for novel phenomena with 100% concentration. In fact it might be better to have a machine that can't be distracted from it's job of looking for things to be distracted by.
Some discoveries. Very slowly. With very limited flexibility and substantially reduced spinoff benefits. Robots are a great way to explore some things but they are not the preferred way to explore everything.
Preferred by who? The scientists? The tax payers? The explorers?
How digestible the plant is by people in space to start with. How the plant interacts with humans in a different environment for another. You cannot discover a lot of things that relate to humans without a human being present.
When it comes time to actually eat a plant grown on mars, then sure we will need a person to do that, but even then it will first be analyzed by machines to make sure it has only safe chemical compounds.
Who said anything about making something deficient? Strawman argument you have there my friend. Don't try to put words in my mouth.
A robot that can't do something is deficient of that ability. By having a mission where humans rather than machines have the responsibility to accomplish certain tasks, you are intentionally making those machines deficient of those abilities. I am saying that as an engineer (someone interested in engineering), the possibility of making something that can do it's job with less human intervention is more exciting than if it requires humans next to it to make it work right. I am disputing your claim that manned missions generate more interest in science and engineering. I think they generate more interest in being an astronaut.
Of course we'll invent good spinoff technologies. But they will be DIFFERENT technologies. There are some technologies that will only and can only be developed if you plan to send people.
And you know exactly which spinoff technologies will be invented, and therefore know that the spinoff technologies of manned mars missions will be more useful than the spinoff technologies of unmanned mars missions?
And this robot vs human thing is a false dilemma. It does not have to be an either/or proposition. We can and should do both.
We were never doing *only* manned missions. There were always machines doing most of the work in space. The difference is that now *all* the space work can be done by machines. Or another way to look at it is that now *all* the human work can be done on the ground.
There already is a tangible benefit to sending people. I outlined a large number of them. It will also take us decades to get to the point where sending people is a realistic option.
The tangible benefits you listed were all related to learning about how to send people to mars. That's like saying the tangible benefit to sky diving is learning how to sky dive. Fine maybe learnign how to sky dive is a benefit. What I am saying is that we should learn to sky dive when it benefits us to be able to sky dive (apart form the benefit of simply knowing how to sky dive).
If all we cared about were the benefits of spinoff technologies. We could simply invest that money in those technologies and probably do it more efficiently than through this indirect approach.
I never claimed that we should. Not sure who's argument you are responding to but it certainly isn't mine.
I never claimed this was your argument. I think you have straw man sensitivity. This was part of my argument to illustrate when *I* think we should send humans and when I think we shouldn't.
How much more money did NASA spend on the Apollo missions than the mars rover missions? I find it ironic that in your example of a human doing more than a machine they are still doing it in a machine.
Sending a big rover that can drive longer ranges is just a matter of spending more money on bigger rockets and more fuel to deliver more cargo into space.
We couldn't go anywhere in space without the precision of computer control systems.
What is this team of scientists going to do on mars? Look at things with their eyeballs and touch things with their fingers? No they are going to be using instruments to collect data (whether it's numbers, images, etc). All that stuff can be sent back to earth for analysis.
It's like saying "We *need* human pilots because a little quadcopter I bought at toys'r'us doesn't have the range of a Boeing 787". Even though not only can an autopilot be installed in a 787, but they already actually have far more sophisticated autopilots than some toy quadcopter.
As opposed to something that's on the scene being controlled by someone that's able to adapt to the situation, and take advantage of the unexpected.
For the price of having a team of scientists on mars, we could probably have hundreds of rovers and teams on earth supporting them. Or we could probably just make far better rovers.
What makes a human better than a rover? That a human can walk faster than a rover can roll? That he/she can climb over terrain better? Those are all things that rovers will can/will get better at (if we are willing to spend the money), where humans will never really improve.
Rovers are probably not going to be good at making high level decisions (e.g. how should I conduct this experiment?), but those sorts of decisions don't need low latency. The decisions that do need low latency (low level decisions (e.g. how should I avoid this rock), are already getting to the point to where the are close to being better than humans (especially in an environment that humans aren't used to).
There really is no reason that high level decisions need to be made on mars. And high level decision making is really the only thing that humans will do better than robots for the foreseeable future.
You get the most sophisticated tool we possess on Mars. One that can make discoveries orders of magnitude faster than any other tool we possess.
The most sophisticated tool we possess, discovered a way to make discoveries without leaving earth.
What can a human being learn about botany in space that a human can't learn on earth by controlling a robot botanist?
and countless other subjects not relevant to mechanical robots
Nothing is relevant to robots. They are robots. Everything they do is something that is relevant to humans. Artificial sensors are better at detecting things than human beings (even the things that are only relevant to humans). A robot will know the CO2 level in a room better than a human ever will. That's we we use instruments to measure CO2 levels and don't just ask people how much it feels like there is.
You'll also inspire a lot of people to get into science and engineering - far more than any robot mission ever could.
As an engineer, (and not an astronaut), I think I am far more interested in making the thing that actually goes to mars and does the work, rather than making something that is so deficient that it requires a human being to be in close physical proximity to operate it.
I think we will invent good spinoff technology regardless of whether we send humans or robots. In fact I would say the *best* spinoff technology to come from the space race were the advances in automation.
You know there used to be a time when we planned (and the russians actually did) send manned spy satellites into space. The job of the person on board was to point the spy camera at interesting things to spy on, and also use the on-board weapon systems to shoot downl other spy satellites. Before we actually finished ours, someone (very smart) realized that the future was to send unmanned spy satellites. "How will the machine possibly do as good of a job as a human?" people said. It turns out that those people just failed to understand what was possible through automation.
I'm not saying we shouldn't have people on mars. We should when it benefits us. Automation removes the *need* to put humans on mars to actually do things on mars. We should go when there is a tangible benefit to going.
We shouldn't send people to mars to repair robots. Robots can repair robots. We shouldn't send people on mars to operate instruments. Robots can operate instruments. We shouldn't send people to mars to point cameras. We shouldn't send people to mars to lift heavy shit. We shouldn't send people to mars to push buttons on a computer. These are all reasons we used to send peopel to places they didn't want (but needed) to go.
We should send people to mars when it is better for those people to be on mars than on earth.