Robotic Executioners

[unixknight]

I was following the discussion in the SOTU thread and @Traveler made some comments referencing the idea of a robotic firing squad.  That got me thinking about the use of robots for such purposes in general, and I find myself viscerally against that notion.  I'm also strongly against robots fighting in battle.  

(Funny how this isn't a science fiction discussion.) 

Isaac Asimov's Robot Laws are, to me, a stroke of genius.  The wisest thing ever written in any sci-fi, IMHO.  I don't think robots (or computers) should ever, EVER be used in the taking of human life under any circumstances.  Why?

Because killing would become too easy, that's why.

As humans, we have a moral aversion to the taking of human life.  Even in those times when we may consider it "necessary" such as capital punishment, self defense or war.  I agree that it's sometimes necessary, but not something to celebrate or enjoy.  The killing of a fellow human is a life changing event, and never to be taken lightly.  

Robots being used for the purpose insulates us from the gravity of it.  By sending the robot in to do it, we can absolve ourselves of any personal responsibility and the problem of desensitization rockets to a new level.  

Remember that original Star Trek episode, "A Taste of Armageddon?"  That's the one where Kirk and crew encounter a civilization at war with a neighboring planet, but despite the massive casualties, there was no physical damage to either world.  The war, you see, was being simulated in a computer, and whenever an attack was launched, the computer would calculate the number of deaths that resulted.  Then, that number of people would be voluntarily disintegrated.  This allowed for the war to carry on without destroying the underlying infrastructure of society.  The result was that the war had been going on for centuries, because the people fighting it were so far removed form the reality of what was happening, that they had no incentive to stop.

Is that our future?  We already have robotic drones out there.  What happens when entire armies of robots are meeting on the field?  Sure, one can say it saves lives by keeping living people out of direct combat, but is that really a justification?  Casualties, suffering, loss and destruction of lives are the reality of war.  It's what makes war an undesirable thing.  Shall we throw away our incentive to avoid war so that we can instead dump resources into these sterile, pointless robot wars? 

On a smaller scale, can the same thing happen in our application of capital punishment?  Will it become easier to execute people if no living person has to throw the switch?

[anatess2]

"We hold to the belief that the man who passes the sentence should swing the sword. If you would take a man's life, you owe it to him to look into his eyes and hear his final words. And if you cannot bear to do that, then perhaps the man does not deserve to die."

- Eddard Stark

[NeuroTypical]

8 minutes ago, anatess2 said:

"We hold to the belief that the man who passes the sentence should swing the sword. If you would take a man's life, you owe it to him to look into his eyes and hear his final words. And if you cannot bear to do that, then perhaps the man does not deserve to die."

I can buy that 100%.  And I believe sometimes capital punishment is appropriate.

Last I heard, all the US killing machines out there (other than land mines and tripwire stuff) all have human operators pulling a trigger.  I've heard Israel has really bumped up the tripwire tech, so they'll have a camera pointed at a window, and a gun will shoot at the window if it detects movement.  

[unixknight]

@anatess2 I had that very quote in mind as I was writing that post.

@NeuroTypical Glad to hear that.  I thought some drones were purely robotic.

[Emmanuel Goldstein]

Read these to see where we may end up going.

 

[NeuroTypical]

8 minutes ago, unixknight said:

 

@NeuroTypical Glad to hear that.  I thought some drones were purely robotic.

I don't really follow military tech as much as I used to in my Tom Clancy readin' days, so don't take my word for it.

[unixknight]

As a side note, robots and A.I.s are no longer purely sci-fi concepts, and we, as a species, need to start deciding now how we're going to deal with them.

Personally, I'm a fan of making use of Asimov's Robot Laws:

• First Law: A robot may not injure a human being or, through inaction, allow a human being to come to harm.
• Second Law: A robot must obey orders given it by human beings except where such orders would conflict with the First Law.
• Third Law: A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

But personally, I'd update them to:

• Zeroth Law: An A.I may not interfere with human agency.
• First Law: An A.I. may not injure a human being or, through inaction, allow a human being to come to harm, except where action would conflict with the Zeroth Law.
• Second Law: An A.I. must obey orders given it by human beings except where such orders would conflict with the Zeroth or First Law.
• Third Law: An A.I. must protect its own existence as long as such protection does not conflict with the Zeroth, First or Second Law.

That would remedy one of the conflicts illustrated in the book I, Robot as well as the film adaptation.

 

Edited February 19, 2019 by unixknight

[mikbone]

[Fether]

Other interesting representations of this are found in Enders Game where Ender thinks he is running a simulation and commits genocide without realizing it. And also a short seen in “surrogate” where the soldier is shown sprinting into danger with no concern of death because he wasn’t really there but was just controlling an avatar robot.

[Traveler]

I use to work for the defense department on a number of weapon systems.  We already have robotics built into weapon systems for 40 years or more.  Without going into too much detail - many such systems have what is called a consent switch or trigger that once activated automated (robotic) systems take over and fire weapon systems with surgical precision. 

 

The Traveler

[NeuroTypical]

My company is moving into the quantum computing market.  I get to sit next to people who will design ways to test qubits, so we can have our quantum computers and revolutionize computing the same way the internal combustion engine revolutionized everything.  

I'm just a lowly beancounter, struggling to understand the basic concepts.  A quantum bit is both a 1 and a 0 at the same time?  Makes my head hurt.  But all the smart people say it's true, IBM has a 20 qubit computer.  In a few years, companies and governments and research institutes should be investing around a billion bucks a year in this stuff.  It's got more behind it than cold fusion fiasco. 

First real quantum computer to the table, gets to crack all the worlds' best cryptography.

[Emmanuel Goldstein]

Don't worry. John Conner is here.

[JohnsonJones]

I think the bigger danger with robots being programmed to kill is that it could eventually become where robots are only defined by their programming and the programming either gets a bug or is done badly.  Then we get a bunch of dead humans trying to survive against killer robots.

[NeuroTypical]

This chart needs a "car decides because you keep driving slow in the fast lane, you need to die" line somewhere.  My vote is that line is at 40%.

 

 

 

[Vort]

Does the electronic fuel injection on a $30,000 car really cost $3000? Seems unlikely.

[Guest MormonGator]

I for one, welcome our new robotic overlords. 

[unixknight]

19 minutes ago, Vort said:

Does the electronic fuel injection on a $30,000 car really cost $3000? Seems unlikely.

Easily, though in that graph it's probably more like $1,500, since about half of that was there before in 1970.  I could believe 3k though.

EFI systems consist of:

• The ECU, a computer which takes inputs from sensors, driving habits and variables for fuel economy, performance and emissions and decided how much fuel to inject into the intake manifold
• 1 injector per cylinder, which pulses in a waveform pattern based on instructions from the ECU (In TBI systems, it's a single throttle body with one or two injectors that sit where a carburetor would be)
• High pressure fuel hose (often around 35psi)
• Electric fuel pump, usually (but not always) located inside the fuel tank
• A variety of sensors, including the oxygen sensor, intake air temp sensor, manifold pressure sensor, coolant temp sensor, mass airflow meter, intake air temp sensor, throttle position sensor, etc.
• fuel rails and return line

(I used to be a Master Certified auto tech before I was reborn as a Software engineer)

Edited February 19, 2019 by unixknight

[mikbone]

https://i.imgur.com/Pk4lgUV.gifv

[NeuroTypical]

1 hour ago, Vort said:

Does the electronic fuel injection on a $30,000 car really cost $3000? Seems unlikely.

I'd agree, but the bar chart is noting the inclusion of various big things, not saying that specific bar is made up only of that thing.  Seems more likely when you consider the overall growing pace of electronics in a car.

I mean, I did grab the thing off the internet because I thought it was interesting, but I did grab it from a fairly respectable corner of the internet.

Edited February 19, 2019 by NeuroTypical

[zil]

1 hour ago, MormonGator said:

I for one, welcome our new robotic overlords. 

Fortunately, our current robotic overlords were born in Sacrament Meeting 3 weeks ago, and haven't had time to become corrupted yet:

Pretty sure that's Sailor Tokiwa Matsu in a Wing Sung 601.

[NeuroTypical]

Not quite robotic executioners, but close: Russian Twitterbots have been helping legitimize the Anti-Vax movement and intentionally sowing discord for years now.  I'm not joking, I'm not being ironic, I'm not exaggerating for effect.  (I routinely do all that jargle, but not doing it here.)

https://ajph.aphapublications.org/doi/10.2105/AJPH.2018.304567

Edited February 19, 2019 by NeuroTypical

[Vort]

52 minutes ago, NeuroTypical said:

In a few years, companies and governments and research institutes should be investing around a billion bucks a year in this stuff.  It's got more behind it than cold fusion fiasco.

I've reversed my course on fusion in general. It's all empty promises. It's not just that the engineering is sooooo far beyond present capabilities; there are actual physical reasons to doubt it can be done using any foreseeable technology.

Fusion, like fission (or coal burning, or natural gas), is just a way to produce lots of heat, which in turn gets converted at an efficiency of around 30% into electricity. The promise of fusion is that it's so incredibly effective and cheap on a per-kWh basis that power becomes practically free.

What anyone over 50 who has paid attention to fusion knows is that in our childhood, we were promised fusion in 30 years. Twenty years later, physicists admitted the technical challenges were greater than anticipated, but that NOW fusion was REALLY only about 30 years off. Today, how long until we have fusion? Well...about thirty years, according to this article. An Economic Times article predicts 60 years, while this Guardian article cites an MIT researcher as predicting 15 years. I'll leave it as an exercise to the reader to predict which one I think it most likely. (Hint: It isn't the Guardian article.)

But there's a much more serious problem, and it's not an engineering problem, it's a physics problem. Pretty much 100% of the articles you read about fusion power tell how it's the power source for our sun. This is false, at least in the way they claim. The fusion that they're trying to produce is not that which powers the sun. It can't be. The sun fuses hydrogen at its core, at temperatures and pressures that are so immensely far beyond our capabilities to reproduce that, outside of the blast of an atomic bomb, it cannot be done. We can't fuse raw ¹H (protons) into a helium nucleus, not in any possible sort of controlled manner. Rather, current fusion research concentrates on various isotopes of hydrogen and perhaps helium that might be fused into helium and lithium, and maybe beryllium, with far lower energy requirements to get the nuclei to fuse.

But this is perhaps just a quibble about Fusion Form X in the sun vs. Fusion Form Y that we're trying to develop. A slight misdirection, maybe, but who cares? If it works, it works, am I right? And there's the problem. From a physics perspective, it gets much worse when you actually think about a working fusion reactor.

Consider this: The sun's power output per unit volume is actually extremely low (3.846 x 10²⁶ watts / 1.4 x 10²⁷ m³ = barely over a quarter of a watt per cubic meter). This is just about identical with the power output of a small mammal's body heat. So if we want to build, say, a one giggawatt fusion reactor—large, but not THAT large—and we can count on getting the same volumetric power density as the sun gets (which we can't*), our reactor would have to be four billion cubic meters in volume. That's a power plant 4 kilometers long, 5 kilometers wide, and 200 meters high. And a cubical geometry for the power plant as I suggest above would be much lower yield, so add another order of magnitude.

It won't work.

*The sun produces fusion because its core is so fantastically hot and high pressure. As I mentioned before, we have no possible hope of reproducing such conditions on earth in any kind of controlled, sustained manner. So even using easier-to-fuse isotopes, there is no way we're going to match the sun's efficiency at producing a fusion environment.

Bottom line: Although controlled nuclear fusion is a theoretical possibility for power generation, the engineering challenges and even the basic understanding of the physical processes involved are so far beyond us that it ain't going to happen in 30 years. If we need an alternative (non-petroleum) energy solution in the immediate future—AND WE DO—fusion is a fool's errand. We should be pouring money into molten salt fission reactors. Now there's something real, that could be developed in five years and online in ten, and that we would have the resources to produce for as long as we can reasonably project the human race to last. Uranium from seawater is effectively a RENEWABLE resource that could provide all our current energy needs for millions of years. Human beings haven't even been around for millions of years. Why are we wasting time and effort on fusion, which is about as achievable today as a manned mission to Alpha Centauri, instead of developing a process we've known about since the 1940s and which was only shelved for development because of military and political expediencies?

/rant

[NeuroTypical]

This is what Vort sounds like.

 

[Emmanuel Goldstein]

3 minutes ago, Vort said:

I've reversed my course on fusion in general. It's all empty promises. It's not just that the engineering is sooooo far beyond present capabilities; there are actual physical reasons to doubt it can be done using any foreseeable technology.

Fusion, like fission (or coal burning, or natural gas), is just a way to produce lots of heat, which in turn gets converted at an efficiency of around 30% into electricity. The promise of fusion is that it's so incredibly effective and cheap on a per-kWh basis that power becomes practically free.

What anyone over 50 who has paid attention to fusion knows is that in our childhood, we were promised fusion in 30 years. Twenty years later, physicists admitted the technical challenges were greater than anticipated, but that NOW fusion was REALLY only about 30 years off. Today, how long until we have fusion? Well...about thirty years, according to this article. An Economic Times article predicts 60 years, while this Guardian article cites an MIT researcher as predicting 15 years. I'll leave it as an exercise to the reader to predict which one I think it most likely. (Hint: It isn't the Guardian article.)

But there's a much more serious problem, and it's not an engineering problem, it's a physics problem. Pretty much 100% of the articles you read about fusion power tell how it's the power source for our sun. This is false, at least in the way they claim. The fusion that they're trying to produce is not that which powers the sun. It can't be. The sun fuses hydrogen at its core, at temperatures and pressures that are so immensely far beyond our capabilities to reproduce that, outside of the blast of an atomic bomb, it cannot be done. We can't fuse raw ¹H (protons) into a helium nucleus, not in any possible sort of controlled manner. Rather, current fusion research concentrates on various isotopes of hydrogen and perhaps helium that might be fused into helium and lithium, and maybe beryllium, with far lower energy requirements to get the nuclei to fuse.

But this is perhaps just a quibble about Fusion Form X in the sun vs. Fusion Form Y that we're trying to develop. A slight misdirection, maybe, but who cares? If it works, it works, am I right? And there's the problem. From a physics perspective, it gets much worse when you actually think about a working fusion reactor.

Consider this: The sun's power output per unit volume is actually extremely low (3.846 x 10²⁶ watts / 1.4 x 10²⁷ m³ = barely over a quarter of a watt per cubic meter). This is just about identical with the power output of a small mammal's body heat. So if we want to build, say, a one giggawatt fusion reactor—large, but not THAT large—and we can count on getting the same volumetric power density as the sun gets (which we can't*), our reactor would have to be four billion cubic meters in volume. That's a power plant 4 kilometers long, 5 kilometers wide, and 200 meters high. And a cubical geometry for the power plant as I suggest above would be much lower yield, so add another order of magnitude.

It won't work.

*The sun produces fusion because its core is so fantastically hot and high pressure. As I mentioned before, we have no possible hope of reproducing such conditions on earth in any kind of controlled, sustained manner. So even using easier-to-fuse isotopes, there is no way we're going to match the sun's efficiency at producing a fusion environment.

Bottom line: Although controlled nuclear fusion is a theoretical possibility for power generation, the engineering challenges and even the basic understanding of the physical processes involved are so far beyond us that it ain't going to happen in 30 years. If we need an alternative (non-petroleum) energy solution in the immediate future—AND WE DO—fusion is a fool's errand. We should be pouring money into molten salt fission reactors. Now there's something real, that could be developed in five years and online in ten, and that we would have the resources to produce for as long as we can reasonably project the human race to last. Uranium from seawater is effectively a RENEWABLE resource that could provide all our current energy needs for millions of years. Human beings haven't even been around for millions of years. Why are we wasting time and effort on fusion, which is about as achievable today as a manned mission to Alpha Centauri, instead of developing a process we've known about since the 1940s and which was only shelved for development because of military and political expediencies?

/rant

So the Sun is a Gravity Engine that has fusion as a side effect? 

 

[zil]

10 minutes ago, Vort said:

fission

OK, Fission (by Nemosine):