Jump to content

For Questions That Don't Merit Their Own Thread


Skyler4856

Recommended Posts

Well, benevolent isn't even a requirement here. Faultless (or, to be more specific, better than a bunch of humans working together) is the issue here with most of computer driven solution. And, to a point, an automated (or close to automated, with not a lot of humans involved) system to move people around according to their wishes, can be assimilated to a public collective transportation system, which tends to be cut down in favor of the fallacies of self driven car. So, if you want to reduce human error, reduce the number of humans driving per humans being driven somewhere. Eventually, assist them with some computer and some data, but keep human intuition in the loop.

Also, what if I need to have a low profile and be of the grid in this automated grid ? There's legitimate reasons (witness protection is one, escaping an abusive partner is another, doing investigative journalism another, etc) to want to get of the grid. How your AI can manage that ? She probably won't see the car and think the space is free, so she can move this heavy trucker right into your spot, right ? Since there's no car here ?

Link to comment
Share on other sites

Well, this is where AI is used widely in place of automated. There's no AI in traffic light, there is automation. And some coordination for special effect (such as green waves), but once they're programmed, their code does not change.

Usually, what is called AI, is second generation code, ie code generated by code. Which neural network does (you code the neurons, the neurons code the functions to make it overfly simplified). Traffic light automation are not second generation code. Mostly because you want to be able to precisely anticipate what the traffic lights will do in specific situation, which you can't do with AI (you can guess, estimate but not know for sure how an AI will react before it reacted, and explaining why it reacted this way might be extremely complex).

In situation where you can have accountability issues (such as being sure that there's only one way that have the green light on a specific crossroads) to be able to fix, improve and share responsibilities, you might not want to rely on second generation code. You want to be able to be sure that if A happens, then the code will do B, each and every time. Which you cannot prove formally for AI (does not means it will not happens).

Litigation department will hate all those self driven vehicles.

Link to comment
Share on other sites

33 minutes ago, TheSaint said:

(In an alt-text comment Zach Weinersmith once said that he could probably write an entire book of comics about douchey robots.) 

Issac Asimov was able to give robots unexpected emergent behavior with only three laws.  I wonder if future robot/AI software will have similar issues with traditional, layer-based software development.  Especially if multiple middleware vendors each include routines to further their own interests.

Link to comment
Share on other sites

12 minutes ago, wumpus said:

Issac Asimov was able to give robots unexpected emergent behavior with only three laws.  I wonder if future robot/AI software will have similar issues with traditional, layer-based software development.  Especially if multiple middleware vendors each include routines to further their own interests.

You mean when AI meets self-serving corporate executives meets clueless government regulators meets The Law of Unintended Consequences? Yeah, we're all doomed.

Link to comment
Share on other sites

2 hours ago, Okhin said:

Well, benevolent isn't even a requirement here. Faultless (or, to be more specific, better than a bunch of humans working together) is the issue here with most of computer driven solution. And, to a point, an automated (or close to automated, with not a lot of humans involved) system to move people around according to their wishes, can be assimilated to a public collective transportation system, which tends to be cut down in favor of the fallacies of self driven car. So, if you want to reduce human error, reduce the number of humans driving per humans being driven somewhere. Eventually, assist them with some computer and some data, but keep human intuition in the loop.

Also, what if I need to have a low profile and be of the grid in this automated grid ? There's legitimate reasons (witness protection is one, escaping an abusive partner is another, doing investigative journalism another, etc) to want to get of the grid. How your AI can manage that ? She probably won't see the car and think the space is free, so she can move this heavy trucker right into your spot, right ? Since there's no car here ?

There is no need for an automated traffic system to know your identity or even your route outside the area its responsible for. 
Its probably smartest that it don't know simply because it make it harder to mess with the system various ways. 
This could be abused so many way from the trivial one of giving you high priority too making visiting one mall much faster than another in a way who would be hard to spot.  
Or identify your ex-wife car as an tractor :) 

Add that most automated systems are broken down into parts, you have one system dependent on an intersection, this is not AI driven simply as its verify that an AI is safe in all settings. 
An automated system can have lots of overrides all it need to do is weave cars trough an intersection with maximum efficiency. At low traffic it would just weave, at higher one you will get queues like with stop signals but the system would stop cars some distance from  intersection so first car passed it at speed limit just a bit after last car in other group left it. 

An AI oversee it for managing traffic. Humans oversee it to see that things works and to set policy, you don't want spillover traffic into residential areas minor roads but they might be used in emergencies as an example. 

27 minutes ago, Okhin said:

you mean, like self driven cars being trapped in salt circles ? (that's my favorite one :p)

Yes that was an hilarious one.
One way to troll humans is a bird house who is an replica of an speed trap automated camera :) 
Trolling AI is so much easier. as in the salt circle, now paint lines at the exits in an turnaround so the AI can not exit :targetpro:
This will be far more fun once you have cars driving around without people in them, or how to steal an ferry full of expensive cars. 

 

Link to comment
Share on other sites

I have looking into electron propulsion do you think there is such thing. Im thinking that you can shoot electrons with magnets and lasers to create propulsion. then use magnets to create a negative froce feild so the electrons dont return. since electrons are electricity so an infinite propulsion system. will that work???????

Link to comment
Share on other sites

1 hour ago, Dr. Kerbal said:

I have looking into electron propulsion do you think there is such thing. Im thinking that you can shoot electrons with magnets and lasers to create propulsion. then use magnets to create a negative froce feild so the electrons dont return. since electrons are electricity so an infinite propulsion system. will that work???????

 

The answer is probably no. I am no expert, but if it was that easy we would have done so already.

The main issue I see is low, low, thrust.

To get higher thrust you either need high energy exhaust or high mass exhaust (they are interchangeable) since:

The equation — E = mc2 — means "energy equals mass times the speed of light squared." It shows that energy (E) and mass (m) are interchangeable; they are different forms of the same thing.Mar 30, 2017

 

If your exhaust is weak, don't expect to arrive anywhere soon.

We already have ion propulsion, which has about the same amount of thrust as a piece of paper hitting your hand. It works great over long periods (years).

Edited by Spacescifi
Link to comment
Share on other sites

7 minutes ago, Dr. Kerbal said:

So we just need to shoot more stuff out. Like a lot, of lot, of lot, of lot, of lot stuff out.

 

Or just go project orion (a nuke equals enough energy to propel to orbit if you do it repeatedly off of a pusher plate).

Lots and lots is essentially why space rocketry is what it is.

A lot of anything requires a lot of energy or mass to push, and since people freak out over nukes and rightly so... we are stuck with using lots and lots of mass (propellant) instead.

Link to comment
Share on other sites

44 minutes ago, Dr. Kerbal said:

oh. well. we could use our old nukes' and use them for rockets. one getting rid of nukes' and too good rockets

 

It is simple in concept.... harder to execute.

Like from a physics standpoint the mathematics checks out as yes, this is possible.

But what are you willing to sacrifice? Nukes equal radioactive death cancer air.

There are NTR that are closed cycle that can reduce that, but they also have lower thrust... meaning they neef normal rocket boosters to even reach orbit.

Open cycle NTR could with the proper design reach orbit... but it spews radioactive death air the whole way.

 

Thus why we have not done so. What we can and should do are often different things.

Link to comment
Share on other sites

First of all, the electrons should be stored somehow.
And the only realistic way to store the mass of the electrons compared to the ship mass, is to keep them in atoms. I.e. 1 proton per 1 electron, plus additional neutrons.

But as a proton is ~2000 times heavier than the electron, so this means that by using just electrons you can't use more than 1/2000 of the ship mass as propellant
If accelerate the protons, you don't need the electrons, or you can use them as an electronic beam to heat the reaction mass.

Link to comment
Share on other sites

20 hours ago, Dr. Kerbal said:

I have looking into electron propulsion do you think there is such thing. Im thinking that you can shoot electrons with magnets and lasers to create propulsion. then use magnets to create a negative froce feild so the electrons dont return. since electrons are electricity so an infinite propulsion system. will that work???????

Don't forget that the electrons will be attracted to your ship 'cause Coulomb's law.  By throwing away those electrons, you just made your ship more positive.  This is why ion thrusters have hollow cathodes.

Edited by Entropian
grahmahr
Link to comment
Share on other sites

12 hours ago, kerbiloid said:

First of all, the electrons should be stored somehow.
And the only realistic way to store the mass of the electrons compared to the ship mass, is to keep them in atoms. I.e. 1 proton per 1 electron, plus additional neutrons.

But as a proton is ~2000 times heavier than the electron, so this means that by using just electrons you can't use more than 1/2000 of the ship mass as propellant
If accelerate the protons, you don't need the electrons, or you can use them as an electronic beam to heat the reaction mass.

Yes an ion drive works much better, add an electron gun as int the rear part of an crt monitor next to it to get rid of electron buildup. 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...