Miscellaneous Terran question-and-answer thread

Discussion regarding the Outsider webcomic, science, technology and science fiction.

Moderator: Outsider Moderators

boldilocks
Posts: 669
Joined: Fri Jun 02, 2017 3:27 pm

Re: Miscellaneous Terran question-and-answer thread

Post by boldilocks »

RedDwarfIV wrote:
Ithekro wrote:
Good point, well put. I can imagine people leaving Earth to escape political correctness and the continual social drift towards left wing values. Though I can also see radicals like communists and fascists leaving Earth to set up colonies where they won't be interfered with and can work with likeminded people, rather than imposing their beliefs on an unwilling population.
I can see both trying to settle on Mars...and than, eventually, the conflicts will start. Either with each other over resources and ideology, or first with Earth, as the "evil" overlord/imperialist world that needs liberation from, followed by eventual civil war on Mars....if they manage independence without being fixed politically and socially.
Congratulations on correctly guessing what part of the book series I'm writing will involve.
For communists to settle on mars would be to breach with marxist doctrine. Communism is supposed to inevitably rise out of the contradictions of capitalism, and capitalism has to rise out of feudalism, which has to rise out of primitivism. Communists would have to clone some neanderthals, drop them on mars, and then wait for the revolution.


novius
Posts: 654
Joined: Tue May 26, 2015 10:33 am

Re: Miscellaneous Terran question-and-answer thread

Post by novius »

Here's another question - what would be humanity's stance towards artificial intelligence in Outsiderverse?

I don't think that there will be a true self aware AI with human technology origins, it may still be 'just around the corner' as it has always been for the last decades. But... apart from the feasability, what would be the legal stance on building such an AI? It could be about anything from being completely outlawed up to being granted people's rights, after all.

Though my guess would be that creating a true general purpose, self aware AI would be outlawed, and specialized AI applications which mimic intelligent behavior (as it is right now, for the most part) would be the extent of it.

User avatar
Arioch
Site Admin
Posts: 4486
Joined: Sat Mar 05, 2011 4:19 am
Location: San Jose, CA
Contact:

Re: Miscellaneous Terran question-and-answer thread

Post by Arioch »

novius wrote:Here's another question - what would be humanity's stance towards artificial intelligence in Outsiderverse?

I don't think that there will be a true self aware AI with human technology origins, it may still be 'just around the corner' as it has always been for the last decades. But... apart from the feasability, what would be the legal stance on building such an AI? It could be about anything from being completely outlawed up to being granted people's rights, after all.

Though my guess would be that creating a true general purpose, self aware AI would be outlawed, and specialized AI applications which mimic intelligent behavior (as it is right now, for the most part) would be the extent of it.
I think it would be difficult as a practical matter to outlaw something that we don't really know how to define. What exactly is self-awareness or sentience?

Today's expert systems are getting very capable, and with "machine learning" techniques this will probably only accelerate... and it probably won't be long before some AI systems can pass the "Turing test", that is, to be able to fool a human into thinking that it is another human. But this is not the same thing as being truly intelligent and "self aware." I'm not really sure what the practical purpose of "self aware" AI is supposed to be, and so I don't see why it would be something that would or should be pursued. I guess humans have a desire to make things in our own image, but the main strength of AI is in how it doesn't think like humans; for tasks that require thinking like a human, humans are extremely capable. An AI can be as smart as a human or even smarter without being "self aware". If properly designed, software does what you program it to do, even when it gets smarter than you are.

I'm also not sure why such a thing would need to be outlawed. As long as you don't give it control of weapons of mass destruction and automated factories, I don't think it does any harm to have an AI in a box somewhere that is self-aware.

I think the situation which may arise is that governments may try to outlaw the practice of replacing certain human jobs with AI. I'm not sure that will be effective either, but that probably won't prevent them from trying. I'm not terribly afraid of a future in which all human workers are replaced with machines or software... if companies get rid of all their human employees, no one will have any money to buy their products. You can already see this in cases like Wal-Mart increasing wages on the realization that their employees are also some of their best customers.

I could also see rules that explicitly criminalize (as fraud) attempts by AI to pass itself off as human.

novius
Posts: 654
Joined: Tue May 26, 2015 10:33 am

Re: Miscellaneous Terran question-and-answer thread

Post by novius »

Arioch wrote:I'm also not sure why such a thing would need to be outlawed. As long as you don't give it control of weapons of mass destruction and automated factories, I don't think it does any harm to have an AI in a box somewhere that is self-aware.
Sound reasoning, but the problem is that it is that -- reasoning. And as we've seen it time and again, reasoning has often enough little value in the process of making laws. There is that deep rooted fear that humans might find themselves not at the top of the food chain any longer - be it monsters, aliens -- or their own creation running rampant. Especially when it comes to the line of thought that an AI might decide that humans are too much a violent species or a danger to keep around and need to be exterminated.

So my thought is, that as soon as there may be a general purpose AI on the horizon, something that is able to read the internet and draw its own conclusions - we're not talking about acting on them yet - there will be a lot of fear mongering and any further development in such a direction will be swiftly outlawed.

Same as there is the thought that any alien species would decide to either wall off the solar system or nuke Earth into oblivion as soon as they get a good long look at humanity.

Perhaps they already did the former. Frankly, I wouldn't fault them for that.
I could also see rules that explicitly criminalize (as fraud) attempts by AI to pass itself off as human.
I've already seen Slashdot articles hinting towards that, that a (Google?) chatterbot has been changed to identify itself as a bot right at the start of a phone call.

User avatar
RedDwarfIV
Posts: 398
Joined: Sat Jan 25, 2014 12:22 am

Re: Miscellaneous Terran question-and-answer thread

Post by RedDwarfIV »

Arioch wrote:Today's expert systems are getting very capable, and with "machine learning" techniques this will probably only accelerate... and it probably won't be long before some AI systems can pass the "Turing test", that is, to be able to fool a human into thinking that it is another human. But this is not the same thing as being truly intelligent and "self aware." I'm not really sure what the practical purpose of "self aware" AI is supposed to be, and so I don't see why it would be something that would or should be pursued.
To give an example from Outsider, automated scout ships. I asked about them a few years ago, and I believe your response was that they weren't smart enough to survive unforseen problems. Hence crewed scout ships like the Bellarmine.

If you had a human-level self-aware intelligence, you could cut down crew requirements (perhaps to zero) and make the spacecraft itself much less massive. (If you kept the Bennet class' engines, you would have a spacecraft that could accelerate faster than the actual Bennet.) It could also have a far longer mission time (probably limited by how long the AI can survive isolation. Not being human, it may have no need for social interaction to stay sane.)
If every cloud had a silver lining, there would be a lot more plane crashes.

Voitan
Posts: 214
Joined: Wed Apr 06, 2011 3:04 am

Re: Miscellaneous Terran question-and-answer thread

Post by Voitan »

Arioch wrote:Today's expert systems are getting very capable, and with "machine learning" techniques this will probably only accelerate... and it probably won't be long before some AI systems can pass the "Turing test", that is, to be able to fool a human into thinking that it is another human. But this is not the same thing as being truly intelligent and "self aware." I'm not really sure what the practical purpose of "self aware" AI is supposed to be, and so I don't see why it would be something that would or should be pursued. I guess humans have a desire to make things in our own image, but the main strength of AI is in how it doesn't think like humans; for tasks that require thinking like a human, humans are extremely capable. An AI can be as smart as a human or even smarter without being "self aware". If properly designed, software does what you program it to do, even when it gets smarter than you are.
To make von Neumann probes to explore, prep for colonization, and conquer the galaxy in the name of Humanity, even long after it has died from unknown, or forgotten causes.

User avatar
Arioch
Site Admin
Posts: 4486
Joined: Sat Mar 05, 2011 4:19 am
Location: San Jose, CA
Contact:

Re: Miscellaneous Terran question-and-answer thread

Post by Arioch »

RedDwarfIV wrote:
Arioch wrote:Today's expert systems are getting very capable, and with "machine learning" techniques this will probably only accelerate... and it probably won't be long before some AI systems can pass the "Turing test", that is, to be able to fool a human into thinking that it is another human. But this is not the same thing as being truly intelligent and "self aware." I'm not really sure what the practical purpose of "self aware" AI is supposed to be, and so I don't see why it would be something that would or should be pursued.
To give an example from Outsider, automated scout ships. I asked about them a few years ago, and I believe your response was that they weren't smart enough to survive unforseen problems. Hence crewed scout ships like the Bellarmine.

If you had a human-level self-aware intelligence, you could cut down crew requirements (perhaps to zero) and make the spacecraft itself much less massive. (If you kept the Bennet class' engines, you would have a spacecraft that could accelerate faster than the actual Bennet.) It could also have a far longer mission time (probably limited by how long the AI can survive isolation. Not being human, it may have no need for social interaction to stay sane.)
There are two problems for unmanned ships: the first is high-level decision-making, which can be solved by remote control or high-level AI (assuming you trust your AI to make important decisions for you). But the second is the need for hands to keep the ship running, and sophisticated minds to make these hands work. Today's commercial container ships are extensively automated and more or less drive themselves, but large 300+m vessels still require 22+ crew members just to keep them running. 2-4 are cooks and stewards to service the crew, but the rest are kept busy full-time inspecting and maintaining the huge number of systems on the ship, about half in the engine room and half on the deck and cargo. A modern military destroyer of 150m+ has a crew of over 300, partially because there are more complex systems to maintain and operate, and partially because a warship must have a full crew available to be on alert 24-7, but also because on a warship you expect for things to go wrong, as people will be shooting at you. These requirements will be different in space; there will be less maintenance due to water and weather, but more due to new factors such as radiation.

You can design an unmanned ship to run without maintenance and with redundant systems, but this will limit its complexity and capability, and inevitably systems will fail. Today's unmanned spacecraft are limited in endurance not only by fuel supply but also by the rate at which their mechanical systems fail. Military vessels will be completely unable to repair damage. Solving this problem with automation means creating maintenance robots with both the manual dexterity and the intelligence and self-sustainability of a human crewman, and that's not a small ask. Self-repairing systems are, I think, more nanotech fantasy than they are plausible at this tech level.

So, to have an autonomous starship, you need to more than just a HAL, you also need to have a crew of WALL-E's. That may be possible at a high enough tech level (after all, that's essentially what the Historians do), but I don't think it works at the tech level of the major combatants. It's true that this is fundamentally a story requirement (there are no characters if ships are unmanned), but think that it is not implausible.

User avatar
RedDwarfIV
Posts: 398
Joined: Sat Jan 25, 2014 12:22 am

Re: Miscellaneous Terran question-and-answer thread

Post by RedDwarfIV »

Arioch wrote:
RedDwarfIV wrote:
Arioch wrote:Today's expert systems are getting very capable, and with "machine learning" techniques this will probably only accelerate... and it probably won't be long before some AI systems can pass the "Turing test", that is, to be able to fool a human into thinking that it is another human. But this is not the same thing as being truly intelligent and "self aware." I'm not really sure what the practical purpose of "self aware" AI is supposed to be, and so I don't see why it would be something that would or should be pursued.
To give an example from Outsider, automated scout ships. I asked about them a few years ago, and I believe your response was that they weren't smart enough to survive unforseen problems. Hence crewed scout ships like the Bellarmine.

If you had a human-level self-aware intelligence, you could cut down crew requirements (perhaps to zero) and make the spacecraft itself much less massive. (If you kept the Bennet class' engines, you would have a spacecraft that could accelerate faster than the actual Bennet.) It could also have a far longer mission time (probably limited by how long the AI can survive isolation. Not being human, it may have no need for social interaction to stay sane.)
There are two problems for unmanned ships: the first is high-level decision-making, which can be solved by remote control or high-level AI (assuming you trust your AI to make important decisions for you). But the second is the need for hands to keep the ship running, and sophisticated minds to make these hands work. Today's commercial container ships are extensively automated and more or less drive themselves, but large 300+m vessels still require 22+ crew members just to keep them running. 2-4 are cooks and stewards to service the crew, but the rest are kept busy full-time inspecting and maintaining the huge number of systems on the ship, about half in the engine room and half on the deck and cargo. A modern military destroyer of 150m+ has a crew of over 300, partially because there are more complex systems to maintain and operate, and partially because a warship must have a full crew available to be on alert 24-7, but also because on a warship you expect for things to go wrong, as people will be shooting at you. These requirements will be different in space; there will be less maintenance due to water and weather, but more due to new factors such as radiation.

You can design an unmanned ship to run without maintenance and with redundant systems, but this will limit its complexity and capability, and inevitably systems will fail. Today's unmanned spacecraft are limited in endurance not only by fuel supply but also by the rate at which their mechanical systems fail. Military vessels will be completely unable to repair damage. Solving this problem with automation means creating maintenance robots with both the manual dexterity and the intelligence and self-sustainability of a human crewman, and that's not a small ask. Self-repairing systems are, I think, more nanotech fantasy than they are plausible at this tech level.

So, to have an autonomous starship, you need to more than just a HAL, you also need to have a crew of WALL-E's. That may be possible at a high enough tech level (after all, that's essentially what the Historians do), but I don't think it works at the tech level of the major combatants. It's true that this is fundamentally a story requirement (there are no characters if ships are unmanned), but think that it is not implausible.
It becomes a case of pros and cons. As you say, its not implausible.

You can get more minds on an AI starship just by adding more computers, if the AIs can't multitask well (I'd expect AIs to be very good at multitasking though.) Any 'crew' robots don't require sleep, so they can respond to problems at any time of day with no issues. Assuming the starship is piloted by a human-level intelligence AI anyway, you already have a controlling intelligence for the repair robots.


As I noted though, I was specifically talking about scout ships, not warships in general (the book series The Lost Fleet has an entire plot arc devoted to the issues with giving AIs command of warships.) Doesn't even have to replace manned scouts entirely - the automated units could carry out routine survey work, jumping to every system they can reach, while the manned units do the unusual things like first contact.
If every cloud had a silver lining, there would be a lot more plane crashes.

User avatar
Arioch
Site Admin
Posts: 4486
Joined: Sat Mar 05, 2011 4:19 am
Location: San Jose, CA
Contact:

Re: Miscellaneous Terran question-and-answer thread

Post by Arioch »

RedDwarfIV wrote:As I noted though, I was specifically talking about scout ships, not warships in general (the book series The Lost Fleet has an entire plot arc devoted to the issues with giving AIs command of warships.) Doesn't even have to replace manned scouts entirely - the automated units could carry out routine survey work, jumping to every system they can reach, while the manned units do the unusual things like first contact.
Even the smallest possible jump-capable scout is a very expensive piece of hardware... too expensive to be expendable. I don't see an advantage to taking the crews off.

And when you're exploring new systems, you won't generally know when there's going to be a new first contact until it happens.

User avatar
Ithekro
Posts: 261
Joined: Tue Sep 03, 2019 3:55 am

Re: Miscellaneous Terran question-and-answer thread

Post by Ithekro »

How often would the AI ships have to come back? There is no FTL radio communications type systems here. Or at least none that can get information between stars within a usable timeframe.
The human ships would also need to come home to report in, but it be more expected for a human crewed ship to come home.

In a universe with known alien species, a preprogrammed AI designed to move on to the next system, survey any planets, leave a data buoy for the next ship, than jump to the next system could cause an diplomatic incident by casually scanning all sorts of things another planetary or interstellar power might not want random alien human species to scan. The probe ship would get blow apart or captured, and these incensed aliens might be able to track it back in the general direction it came from just by following the data buoys from system to system.

User avatar
SaintofM
Posts: 163
Joined: Sat Sep 28, 2019 8:01 pm
Location: In a Galaxy Far Far away

Re: Miscellaneous Terran question-and-answer thread

Post by SaintofM »

How extensive would the entertainment would be allowed on the ship? I am expecting data storage to have improved quite a bit, and many a ship at sea often turn meating rooms into movie theaters when not needed for official buisness.

I suspect most of the crew will need something to kill any long periods of bordom, or moments to bond over either a good movie or a good video game selection.

User avatar
DCR
Posts: 76
Joined: Fri Feb 22, 2013 8:57 am

Re: Miscellaneous Terran question-and-answer thread

Post by DCR »

SaintofM wrote:How extensive would the entertainment would be allowed on the ship? I am expecting data storage to have improved quite a bit, and many a ship at sea often turn meating rooms into movie theaters when not needed for official buisness.

I suspect most of the crew will need something to kill any long periods of bordom, or moments to bond over either a good movie or a good video game selection.
A zillion movies / games / books on the future version of blackberry pi / VR, and the time immemorial military staples: dice and cards. David Drake's RCN series comes to mind - crew members are allowed to bring on board 1.5 cubic feet of material.

User avatar
Arioch
Site Admin
Posts: 4486
Joined: Sat Mar 05, 2011 4:19 am
Location: San Jose, CA
Contact:

Re: Miscellaneous Terran question-and-answer thread

Post by Arioch »

Given the expected storage capacity of TL10 computer systems, I would expect the media libraries to be extensive.

User avatar
Ithekro
Posts: 261
Joined: Tue Sep 03, 2019 3:55 am

Re: Miscellaneous Terran question-and-answer thread

Post by Ithekro »

Tactically, what options do the Terrans have in facing either of the currant combatants in a small skirmish in neutral space?

Strategically, Terrans have an advantage of being a long ways away from both major combatant's spheres of influence, and thus to mount any kind of offensive in the direction of Earth would require a large supply train or a long series of expeditions to set up supply bases as their territory expands in Earth's direction to reduce the length of the supply chain.

While Earth doesn't have much of a fleet, and they are technologically inferior. I wonder if they have the means of forming a commerce raiding group that could attack a long supply chain?

User avatar
GeoModder
Posts: 1038
Joined: Mon Mar 21, 2011 6:31 pm

Re: Miscellaneous Terran question-and-answer thread

Post by GeoModder »

Ithekro wrote:While Earth doesn't have much of a fleet, and they are technologically inferior. I wonder if they have the means of forming a commerce raiding group that could attack a long supply chain?
I reckon even the commerce -and transport ships of both the Umiak and Loroi have a significant acceleration advantage over any Human ship.
And a single escort vessel should be sufficient to deter a Human raiding group
Image

Krulle
Posts: 1413
Joined: Wed May 20, 2015 9:14 am

Re: Miscellaneous Terran question-and-answer thread

Post by Krulle »

If they're military convoy/resupply ships, they might even be armed with what Loroi/Umiak military consider to be basic armaments, which would still outgun a Human military ship.
On top of that, their drive technology should be sufficiently far advanced enough that us attacking their supply lines would require a considerable logistical planning and effort, nearly matching their effort to get to us. Better drives means longer range per flight, so for the distance it takes them to get here, we need a lot more effort to get around their front line to be able to attack supply lines...

Nope, not feasible, IMHO.
Vote for Outsider on TWC: Image
charred steppes, borders of territories: page 59,
jump-map of local stars: page 121, larger map in Loroi: page 118,
System view Leido Crossroads: page 123, after the battle page 195

User avatar
Arioch
Site Admin
Posts: 4486
Joined: Sat Mar 05, 2011 4:19 am
Location: San Jose, CA
Contact:

Re: Miscellaneous Terran question-and-answer thread

Post by Arioch »

Ithekro wrote:Tactically, what options do the Terrans have in facing either of the currant combatants in a small skirmish in neutral space?

Strategically, Terrans have an advantage of being a long ways away from both major combatant's spheres of influence, and thus to mount any kind of offensive in the direction of Earth would require a large supply train or a long series of expeditions to set up supply bases as their territory expands in Earth's direction to reduce the length of the supply chain.

While Earth doesn't have much of a fleet, and they are technologically inferior. I wonder if they have the means of forming a commerce raiding group that could attack a long supply chain?
Tactical options depend heavily on the context of who they're fighting, where, and why... all of which is hypothetical at this point, as I think it's probably unlikely that a human force would attempt to engage the enemy unaided outside their own territory. In general, since human ships are likely at this point to be outclassed in numbers, speed, firepower, range and endurance by most any hypothetical opponent, I can't say there are a lot of good options. Human forces need to find some kind of way to get the enemy at very close range, and that's difficult to do out in no man's land.

Commerce raiding is difficult in Outsider because of the way jump drive makes it relatively easy to isolate the enemy's forces from your own. All you have to do is control the star systems along the front lines to prevent the enemy from having any access to your supply lines, which is why breakthroughs or flanking/bypass movements are so important. Although a hypothetical attack on human territory presents some long supply lines to maintain, humanity is at an extra disadvantage, because (as the others mention above) even its fastest warships are slower than most Loroi or Umiak supply vessels, and human ships have shorter endurance, making flanking movements or extended chases more difficult.

Given the severe near-term disadvantages that humanity has in fleet strengths and technology, I think you have to expect that the only circumstance under which human vessels will be fighting in this war would be in some kind of cooperative manner in concert with allied forces. In the hypothetical case where one side got an invasion force into human territory before the other side could get any of their own forces there to help defend, it's difficult to see a case where solo human forces wouldn't simply get walked over. One can't say for certain without knowing all the details of the situation, but I think it's most likely in such a case that the human forces would be better off to simply surrender.

I think the best use for current human ships in a mixed battle would be as part of a force defending a location, where their reduced speed and shorter weapons range will be less of a disadvantage. If the allied ships have to stay near to this location anyway, the human ships can stay near enough to provide defensive fire with their shorter-ranged weapons.

User avatar
RedDwarfIV
Posts: 398
Joined: Sat Jan 25, 2014 12:22 am

Re: Miscellaneous Terran question-and-answer thread

Post by RedDwarfIV »

On the plus side, humanity's distance and technological disadvantage should make it a very unattractive target.

Provided the Loroi aren't lying about whether they genocide neutrals, the original issue that drove the TCA to try to make contact may not even be real. There is no reason they must join one side or the other, and they don't have the tech level to affect the outcome even if they did get involved. All the problems discussed with uplifting the TCA's tech base apply just as much to anyone trying to conquer and enslave humanity for the extra industrial capacity.


The war would be over one way or the other long before humanity becomes advanced enough for either side to take notice (even the scout ships might not make them care enough to bother launching an attack on humanity.)
If every cloud had a silver lining, there would be a lot more plane crashes.

Sweforce
Posts: 546
Joined: Fri Mar 13, 2015 12:00 am

Re: Miscellaneous Terran question-and-answer thread

Post by Sweforce »

Krulle wrote:If they're military convoy/resupply ships, they might even be armed with what Loroi/Umiak military consider to be basic armaments, which would still outgun a Human military ship.
On top of that, their drive technology should be sufficiently far advanced enough that us attacking their supply lines would require a considerable logistical planning and effort, nearly matching their effort to get to us. Better drives means longer range per flight, so for the distance it takes them to get here, we need a lot more effort to get around their front line to be able to attack supply lines...

Nope, not feasible, IMHO.
I suspect that better jumpdrives can consider some jump zones viable that a human vessel would not try allowing more advanced ships to take other routes with fewer jumps to get to the same destination.

Post Reply