-
#140
by
Marsbug
on 14 Jan, 2010 16:33
-
That's an invalid comparison.
Thats what I was trying to say. My line of reasoning was that the state of robotics has moved on so far from then. What was your line of reasoning?
The question at hand is whether HSF can be justified by the science done with manned missions.
As far as planetary science goes: I've no idea if it can be, but without an example of human exploration of another body (if we're not counting the moon) I have no idea how to approach the question in any case!
Same with AI ethics- without an AI to ask how can we know?
We're veering way off topic methinks.
-
#141
by
khallow
on 14 Jan, 2010 16:37
-
True. Especially for Mars. A human Mars surface mission would be magnitudes more worthy than current robotic missions are. But we don't have the money.
Also, at some point in the future, advancements in robotics, AI and also band-stream may just result in robotic missions which combine a geologist's ability to identify interesting targets and features with the comparatively low costs of a robotic mission and its other positive features (longevity, no risk of a life etc.). Think about the possibility of a robot sending back a 3D image of its nearest environment in immense detail in which geologist can move around on Earth in VR and select further targets.
The geologist still can't sample or interact with the Mars environment in real time. My view is that the fundamental problem with current space science is that effectively, there are a bunch of niches that can be filled by a single unmanned mission. For example, if there's a Venus mission going on, then it's very unlikely that another Venus mission will be approved.
A manned mission (or series of missions) would fill the niche with a mission ten to one hundred times the cost of the minimal needed. That's inefficient from a funding standpoint. You can fill more niches by dropping manned missions even if the science becomes vastly less productive as a result.
-
#142
by
Lambda-4
on 14 Jan, 2010 16:37
-
1. First, we're a long ways from making that happen.
2. Second, there are a lot of ethical issues surrounding the use of human level or greater AIs. What do you do with the AI after the mission is finished? Just terminating a human level intelligence appears to me to be crudely equivalent to terminating the existence of a human.
3. There's also the matter of restricting the capabilities of the robot so that you can't create a situation that quickly balloons out of control.[...]
4. My view is that even for expeditions that use highly advanced AI comparable or better than human level intelligence, it'll be a good idea to have someone local in the loop to insure against things getting out of hand. Even if robots can do it all, someone will want a human middle manager in there to supervise.
1. I disagree. We are at the stage where robotics become more semi-automatic right now. Not in the sense you are thinking of, but in a mission critical sense. A robot that has the capability to path its own way on Mars to a distant destination and does an analysis (taking sample, putting sample in analyzer, relaying back data) is reality now. And that's what we are talking about. Humans did the same thing on the Moon 90% of the time, they had scripted missions (points where to go to, sample objectives etc.) that they were carrying out with communication and back-up from Earth.
2. We are not that far and I doubt we will get that far anytime soon. "AI" is used not only in relation to real intelligence that includes creative thinking and own problem solving, but also relates to more complex scripting of certain tasks. There is no ethical question here, that a Mars rover can identify obstacles in its way and move around them to get to its destination without a human saying so doesn't equate to a rover being alive.
3. That's more or less sci-fi and not applicable for this discussion.
4. Disagree, at least with regard to the surface missions that we will see in the next decades. Earth-based control with a time-lack is sufficient supervision.
-
#143
by
Lambda-4
on 14 Jan, 2010 16:43
-
What was your line of reasoning?
I probably misread your post. My point was that a valid comparison of human exploration and robotic exploration takes all robotic missions and their costs on the one side and compares it with the science done by human exploration and their costs. Which side has advanced science further per (billion) dollar invested wins the comparison.
As far as planetary science goes: I've no idea if it can be, but without an example of human exploration of another body (if we're not counting the moon) I have no idea how to approach the question in any case!
As specified above, we can of course compare human exploration and robotic exploration of planetary bodies. With the exception of the Moon where the science return in total of HSF has been higher for lunar geology, robotic exploration wins by default for every single object ever examined in our Solar System - including Earth.
We're veering way off topic methinks.
Agreed. Last post on this subject from my side.
-
#144
by
TrueBlueWitt
on 14 Jan, 2010 16:47
-
True. Especially for Mars. A human Mars surface mission would be magnitudes more worthy than current robotic missions are. But we don't have the money.
Also, at some point in the future, advancements in robotics, AI and also band-stream may just result in robotic missions which combine a geologist's ability to identify interesting targets and features with the comparatively low costs of a robotic mission and its other positive features (longevity, no risk of a life etc.). Think about the possibility of a robot sending back a 3D image of its nearest environment in immense detail in which geologist can move around on Earth in VR and select further targets.
The geologist still can't sample or interact with the Mars environment in real time. My view is that the fundamental problem with current space science is that effectively, there are a bunch of niches that can be filled by a single unmanned mission. For example, if there's a Venus mission going on, then it's very unlikely that another Venus mission will be approved.
A manned mission (or series of missions) would fill the niche with a mission ten to one hundred times the cost of the minimal needed. That's inefficient from a funding standpoint. You can fill more niches by dropping manned missions even if the science becomes vastly less productive as a result.
Why can't a geologist interact in real time if he's on the surface? I agree having him in orbit teleoperating a robot may be less useful.. though perhaps more cost effective.
The rest of your argument has valid points.. Robotics is cheaper.. without a doubt.
However, does your argument count on the fact congress/WH will be willing to fund NASA at the same level for strictly robotic missions as it will when manned missions are involved?
-
#145
by
khallow
on 14 Jan, 2010 16:48
-
1. First, we're a long ways from making that happen.
1. I disagree. We are at the stage where robotics become more semi-automatic right now. Not in the sense you are thinking of, but in a mission critical sense. A robot that has the capability to path its own way on Mars to a distant destination and does an analysis (taking sample, putting sample in analyzer, relaying back data) is reality now. And that's what we are talking about. Humans did the same thing on the Moon 90% of the time, they had scripted missions (points where to go to, sample objectives etc.) that they were carrying out with communication and back-up from Earth.
Marsbug said "with levels of decision making autonomy, and creative on-the-spot-thinking, comparable to a human". That's far beyond what you are referring to (for example, the "10%" of human decisions that you ignored). Points 2-4 were made given that. If robots never reach human-level decision making capability (in our lifetimes), then those points are not an issue.
-
#146
by
Lambda-4
on 14 Jan, 2010 16:54
-
If robots never reach human-level decision making capability (in our lifetimes), then those points are not an issue.
My point was that it is irrelevant if they come close to human creativity or not. To come close to the efficiency of a human explorer we don't need that kind of AI, we just need a better AI than we have now, improved in the areas where we already have semi-autonomous robotic activity.
-
#147
by
Robotbeat
on 19 Jan, 2010 01:43
-
Even with advances in artificial intelligence, there may be things--potentially very important things--that robots would miss but a human would notice.
A humorous fictional example is expressed in the children's movie "Planet 51." An astronaut lands on a planet which he thinks was uninhabited because the AI rover sent beforehand only wanted to take pictures of rocks (it's inhabited by aliens that look like Sea Monkeys but act like 1950s Americans). Obviously, a mistake that obvious would never happen, but something much more subtle could. It's possible the most interesting things yet to be discovered in space are not something that you could train an AI to do (at least not for a very long time). Human beings, on the other hand, have the advantage of being honed by evolution over millions and millions of years to find interesting things in diverse environments.
Besides, the most compelling argument isn't the science output: Humankind will clearly remain limited to Earth unless we step out into space.
To modify a Reagan quote:
"We are too great a species to limit ourselves to small dreams."
Also, on the subject of Flexible Path, I think it is a good path forward, especially going to Phobos. Phobos WILL inspire, and I think would surely lead to a Mars landing soon after.
If America is in terminal decline (I don't think it is), then someone will take our place, and will inspire people around the world with their space exploration.
-
#148
by
Lampyridae
on 19 Jan, 2010 02:17
-
A point worth mentioning is that, as much as robotics is advancing, so too is medicine and biotechnology. If you could put your astros into hibernation for the trip, then even missions to Mars could become a lot more affordable. Couple that with advanced closed-loop life support and that's big savings. Advances in suit developments will also make EVA a lot easier.
As for America being in decline, it's not really a decline as a reality of industrial development. Once you replace outhouses with toilets, telephones with the internet, there's not a lot more room for development. Space is the next big frontier but it's the most difficult one to get money out of. Still need that magic $200/kilo.
-
#149
by
Ben the Space Brit
on 19 Jan, 2010 12:38
-
Just out of interest, does anyone know the launch window dates for NEOs (within the 180-day mission limit mentioned during the Augustine Commission hearings) from 2020 onwards? If you could direct me to a link where I can see or even calculate these for any solar system object, that would be gratefully appreciated.
-
#150
by
William Barton
on 19 Jan, 2010 12:57
-
I'll ask this question again (and as often as necessary), since no one seems to offer an answer that doesn't boil down to sociopolitical gobbledegook. How do you quantify the value of the "scientic return" from a space mission? Until you can do that, you can't actually determine the cost coefficient for manned vs. unmanned space exploration. Therefore, the entire discussion is political, and it's content is utterly irrelevant.
My personal opinion is, without an intent to send a manned mission at some point in the future, unmanned space exploration has a value of zero. Which is, indeed, sociopolitical. I know of no scientific or economic argument that justifies any space activity whatsoever much beyond GEO. By contrast, I can muster any number of sociopolitical arguments for both manned and unmanned space exploration.
-
#151
by
robertross
on 19 Jan, 2010 13:16
-
I'll ask this question again (and as often as necessary), since no one seems to offer an answer that doesn't boil down to sociopolitical gobbledegook. How do you quantify the value of the "scientic return" from a space mission? Until you can do that, you can't actually determine the cost coefficient for manned vs. unmanned space exploration. Therefore, the entire discussion is political, and it's content is utterly irrelevant.
My personal opinion is, without an intent to send a manned mission at some point in the future, unmanned space exploration has a value of zero. Which is, indeed, sociopolitical. I know of no scientific or economic argument that justifies any space activity whatsoever much beyond GEO. By contrast, I can muster any number of sociopolitical arguments for both manned and unmanned space exploration.
Well, to me we need a Earth-based unit of measure. How do we measure the value of discovery in the Antarctic, or deep sea submersibles, or geologic digs in Egypt?
In the end, there is no tangible way to measure it, it can be argued it is simply the knowledge gained. Now, you could get lucky along the way, find something so unique from the oceans that it provides a cure for cancer, or a snail that has a special steel-layered sheel that could provide better armour for troops, or simply the crowds that will flock to see some new discovery.
Science is the accumulation of wealth in terms of knowledge. It may have short term benefits, long term benefits, unequalled benefots, or no benefit at all. But I would like to know how large a collection of encyclopedia books there would be today for all the kowledge gained through science alone. Good thing we have DVDs and such storage media.
-
#152
by
Lambda-4
on 19 Jan, 2010 14:18
-
Just out of interest, does anyone know the launch window dates for NEOs (within the 180-day mission limit mentioned during the Augustine Commission hearings) from 2020 onwards? If you could direct me to a link where I can see or even calculate these for any solar system object, that would be gratefully appreciated.
The length of NEO missions, their viability and date of launch windows depend on the delta-v you can provide, your requirement of the length of the launch window and length of stay at the NEO and whether you want a prior launch window for a precursor robotic mission. Due to these very specific variables and quite complicated orbital mechanics when it comes to NEO rendezvous calculation (you need precise measurements of the NEO trajectory you want to visit), it takes some time to find a suitable candidate for a certain date and is quite an effort. I am not aware of a comprehensive list that has been established by NASA yet or published by anyone. I know that for their recent Flexible Path internal memo they looked at a variety of NEOs.
-
#153
by
mikegi
on 19 Jan, 2010 15:39
-
... I know of no scientific or economic argument that justifies any space activity whatsoever much beyond GEO ...
Large scale experiments on the validity of General Relativity, etc. If those experiments show our fundamental understanding of physics is incorrect, it would have huge potential to change things in the future.
Edit: it's difficult to estimate the economic impact of changes in fundamental physics. However, it's fairly obvious that the taming of electromagnetism in the 1800s was worth 10s or 100s of trillions of $$$ in the 1900s.
-
#154
by
agman25
on 19 Jan, 2010 15:44
-
I'll ask this question again (and as often as necessary), since no one seems to offer an answer that doesn't boil down to sociopolitical gobbledegook. How do you quantify the value of the "scientic return" from a space mission? Until you can do that, you can't actually determine the cost coefficient for manned vs. unmanned space exploration. Therefore, the entire discussion is political, and it's content is utterly irrelevant.
My personal opinion is, without an intent to send a manned mission at some point in the future, unmanned space exploration has a value of zero. Which is, indeed, sociopolitical. I know of no scientific or economic argument that justifies any space activity whatsoever much beyond GEO. By contrast, I can muster any number of sociopolitical arguments for both manned and unmanned space exploration.
Can you clarify what you mean by scientific return. Does learning about the origins of the universe or life on Earth qualify as return. Or do you mean financial returns.
-
#155
by
MP99
on 19 Jan, 2010 16:02
-
... I know of no scientific or economic argument that justifies any space activity whatsoever much beyond GEO ...
Large scale experiments on the validity of General Relativity, etc. If those experiments show our fundamental understanding of physics is incorrect, it would have huge potential to change things in the future.
Edit: it's difficult to estimate the economic impact of changes in fundamental physics. However, it's fairly obvious that the taming of electromagnetism in the 1800s was worth 10s or 100s of trillions of $$$ in the 1900s.
Is anyone holding off building some fantastic widget based on General Relativity just because it hasn't been properly validated yet?
If they built it, wouldn't that be validation in itself?
cheers, Martin
-
#156
by
khallow
on 19 Jan, 2010 17:18
-
I'll ask this question again (and as often as necessary), since no one seems to offer an answer that doesn't boil down to sociopolitical gobbledegook. How do you quantify the value of the "scientic return" from a space mission? Until you can do that, you can't actually determine the cost coefficient for manned vs. unmanned space exploration. Therefore, the entire discussion is political, and it's content is utterly irrelevant.
My personal opinion is, without an intent to send a manned mission at some point in the future, unmanned space exploration has a value of zero. Which is, indeed, sociopolitical. I know of no scientific or economic argument that justifies any space activity whatsoever much beyond GEO. By contrast, I can muster any number of sociopolitical arguments for both manned and unmanned space exploration.
I'm not going to defend the idea that space science has considerable value, but I will note that in another thread I talked about a
space mining idea that might make sense. It's basically a way to build considerable infrastructure on the Moon and use that infrastructure to mine a big gold find. Ignoring the viability of my approach, it still remains that I have to find this ore body in order to mine it. For purposes of my model, the ore body has a value of $1.2 billion in 2050 dollars. Using the economic time value of money assumptions that I use, that would be a value of $75 million in 2010 dollars. So prospecting on the Moon now could generate $75 million of value by finding this sizable PGM ore body.
In other words, even if most space science is intangible in any benefits it might have, there is the possibility of finding concrete value from space science.
-
#157
by
mikegi
on 19 Jan, 2010 19:48
-
Is anyone holding off building some fantastic widget based on General Relativity just because it hasn't been properly validated yet?
If they built it, wouldn't that be validation in itself?
cheers, Martin
You're looking at it backwards. Invalidation of GR and other current physics theories leads to development of new physics, which could result in completely new devices, energy sources, etc.
-
#158
by
Cons
on 19 Jan, 2010 19:54
-
I'll ask this question again (and as often as necessary), since no one seems to offer an answer that doesn't boil down to sociopolitical gobbledegook. How do you quantify the value of the "scientic return" from a space mission?
First order estimate: Number of peer reviewed publications.
-
#159
by
Robotbeat
on 19 Jan, 2010 20:30
-
I'll ask this question again (and as often as necessary), since no one seems to offer an answer that doesn't boil down to sociopolitical gobbledegook. How do you quantify the value of the "scientic return" from a space mission?
First order estimate: Number of peer reviewed publications.
A little quibble: think of the number of peer reviewed publications on string theory. The possibility remains that it's just a beautiful mathematical construct that has little to do with the physical universe. In that case, its scientific return would be close to zero... perhaps even negative, since a lot of bright people have spent their strongest years working on it rather than other problems in physics. That said, I don't necessarily believe string theory IS necessary just BS, but there are certainly lots of theories that are BS but still generate many peer reviewed publications every year (and countless other non-peer-reviewed publications on arxiv.org...).
But, that is off-topic. In light of the fact that bogus theories (near zero scientific return) can still generate peer-reviewed publications, perhaps a better metric can be found?