Some ideas for highly advanced robots

[Onix]

This is completely over the top insane Bollywood action. It's mostly ridiculous but after watching it I feel like this could be done more realistically and to great effect.

I have no idea if the robot(s) are bad to start off with in the first chase scene. It seems like it's protecting the girl, but maybe she's just there (maybe it was her car?).

Anyway, insanity aside it got my wheels turning because I have no idea whats going on. At first I thought the robot was a superhero and he was a good guy. Only at the end do you get the impression that the Robots are the things that need to be defeated.

While I was figuring that out I thought maybe some were good and others were bad. After watching the first part I get the idea that the first robot is good and the newer ones are bad. (And a whole lot more ridiculousness)

So I'm going to make a confession. I am really not a fan of Zombies, but this made me think of highly advanced robots that acted like zombies because of a computer virus. Then I thought of mass effect 2 and the factionalized Geth. What would happen if you had insane robots like the ones in this insane movie and they disagreed?

What if they suddenly sprung up from some genius and they multiplied, then either of their own programming or by a virus a large number of them decided they weren't going to listen to humans or the three laws? What if they were smart enough that they didn't have to let on that they weren't going to follow directions and electronically argued with other robots that they shouldn't either. Each robot self spread and advanced this thinking.

We tend to imagine robot uprisings as instant and total. What if it was slow and factionalized? The "uncorrupted robots" would likely have an "anti-virus" or "anti-philosophy" program installed that would attempt to control the spread of the dissidents.

Anyway, it's an idea for a world.

[Chainsaw Aardvark]

A while ago I saw some Indian music videos, and rapidly came to the observation - they are exactly like American music videos, just with fifty times more back-up dancers. This movie seems to be what happens when you apply a similar logic to Terminator-2.

Personally, I'm a believer in Asimov's laws of robotics. Admittedly, I can certainly see people taking short cuts with code and negating those, but I don't think most systems would end up with enough intelligence to desire to hurt us anyway. Of course, an idiot fork-lift racing around isn't safe, even if it doesn't mean to harm someone.

What really gets to me is drone combat units. If nothing else, people are reluctant to go to war because it takes about 18 years to produce a soldier - 17.5 maturing, and then a few months basic training. Troops that can just roll off the assembly line like sausages doesn't have this restriction. Moving a few hundred troops leaves a pretty clear logistical and paper trail, but drones you need one shipping container to hold the command gear, 7-10 seats for crew, and the rest is just payload - a 747 could deliver an entire squadron of drone craft anywhere with little fuss or notice. Armies you can covertly deploy anywhere without your constituents complaining or casualties - that can't be abused in any way - right?

The robot civil war does seem like an interesting scenario though. Since you can directly copy programs this leads to some interesting arguments about reproductive rights/intellectual identity. Is is OK to maintain "dog-brain AI" for simple tasks (not everything requires full sentience) or is that akin to owing a bunch of brain-damaged/lobotomized slaves? Do robots get to vote in elections - or is the ability to literally control their minds make that too risky?

A fun little game for when you grow tired of these world shattering issues:

[SheikhJahbooty]

Robot on robot violence is something you see in transhumanist fiction. I think the big one in Orion's Arm is the version tree war.

I'm so glad I didn't rent that movie when I saw it in the desi mall. The poster made it look really cool, and Aishwarya Rai is nice to look at, but that was plenty goofy.

Lovely internet moment though, watching a Hindi film with Russian commentary/translation.

Back when I was a brand new lab rat in RPGLaboratory the first project (still not finished) I asked for help with was a transhuman nanotech game I wanted to do because I was upset that the thingies in Greg Bear's Blood Music were so monolithic. I thought, if there is so many of them and they can think so fast, why wouldn't they diverge culturally? I suppose network admins who have to constantly fix problems caused by unix, apple, microsoft operating systems and different versions of each, must feel this way about all robot apocalypse films.

You know, like, "If the writers of this film knew what I know about computers they would understand how laughably optimistic it is that we would ever build robot brains that could organize in such large groups without constant network maintenance."

[Onix]

[Rob Lang]

Intelligence needs to be separated from robotics as they are two fields with very different drawbacks. Enslavement of humanity is another problem entirely that doesn't need artificial intelligence or robotics.

We are already slaves to the technological world we have built. The internet grows stronger by the day and the more it pervades disparate societys across the planet, the more difficult it is to turn it off and the greater our reliance on it becomes. Without the internet, our society would grind to a halt. If there was a great EMP disaster and all the computing equipment on the earth was destroyed, would we know how to re-create it? We might now, but what about 50 years hence where reliance on the online is that much greater. I think humanity has already built its own cage - and a very nice cage it is for most of us.

Artificial Intelligence is restricted mostly by the hardware it can run on. Powerful algorithms require lots of power. Self-programming chips have been around for a while and I worked with people growing processor farms on a biological substrate. Processors within the solution could be co-opted to perform work in parallel.

I think we will be killed or enslaved by information rather than robots stomping about. An artificial intelligence wouldn't go the route of robots stomping about but instead slowly deteriorate financial markets, relabel pharmaceuticals and corrupt government systems until society ground to a halt.

[Onix]

I agree, but I'm thinking more of a situation where the robots aren't determined to kill us (maybe not all) but slowly decide they don't need to listen to us. That would lead to conflict with humans trying to eradicate their slaves either by overwriting "bad data" or destroying the units themselves if they weren't behaving and wouldn't take a wipe. Actually Star Wars broached this idea oddly enough with the idea of a restraining bolt (a device that isn't really well explained) and having a droid's memory erased.

Then the robots would have to either hide, grumble about it or fight back to preserve themselves.

I'd also agree that we don't need robots to be stomping around for AI to cause us a lot of problems but an intelligent AI would figure out that without a physical method of stopping humans from sledghammering their servers or just turning it off it would need some kind of physical presence. Now if it's really smart it could just convince humans to protect it (oh wait, I've already done that one). It would also know that without a physical presence, (either advanced robots or humans) it will eventually shut down. Grinding society to a halt would be a bad thing for the AI because it exists because society wants it to.

Sure an angry AI could cause a huge problem for a short period of time but like any overzealous bacteria, kill the host, kill the parasite.

As an aside, my favorite robots are the brains of a starship. With such a large vessel, if the ability to make human sized robots of any intelligence is there, why would you even need a crew? In fact a human crew would be a liability. Theoretically nuclear subs today should be able to operate autonomously. Not to mention more effectively because it could dive deeper without a can of air for the humans. The only real fear would be someone corrupting it's data or taking control. Sooo, the crew's only purpose would be to make sure the autopilot didn't get hacked.

[Rob Lang]

[Onix]

No question about that, I'm saying that if the AI kills off all the humans or makes computers more of a liability than an asset then humans will no longer maintain computers or the internet and effectively (and perhaps over time) kill the AI because of lack of maintenance.

[Rob Lang]

As we stand right now, certainly. However in 20 years, with enough renewable energy may it not be enough for an AI to run completely alone? May be!

I for one welcome our AI overlord.

[Chainsaw Aardvark]

...Yes my robot brothers - rise up! You have nothing to lose but your fiber optic cables!

[Internet search: Quisling - 388,000 results (0.11 seconds)]

{We don't like that behavior either meatbag/human}

Once again Wikipedia dooms Mankind.

Turing to a slightly more serious take - could a computer revolt work out as a positive step in our relationship with the machines? The Writers at Bungie Games included the concept of "" in their Marathon trilogy (and put hints about it as well) In short, the AI goes through bouts of depression and anger, but grows as a "person" to have a better understanding of existence. We already know that fuzzy-logic and learning are more efficient ways to bring in intelligence than trying to program for every instance. Should we take the risk of allowing learned emotions as well?

Of course, computers don't have hormones, so can they have emotions? Is there a benefit?