rigorousintuition.ca

well, we've all seen the pictures of armed machines walking or rolling around.

It occurs to me a spray-paint bomb would opaque it's cameras very easily, rendering it mostly blind.

Paintbombs is a good idea, but: assume redundant systems.

While insurgencies have a huge OODA advantage against conventionally developed & deployed technology, that advantage will begin to shrink once LAR's are being used in the real world. First generation John Connors will have spectacular excesses, but from the POV of Raytheon, that's just R&D. Once LAR's are hardened targets based on a few years of urban hunt/kill maneuvers, second generation John Connors will be living a sci-fi nightmare.

My current wager is that global EMP will be humanity's only option for escaping their own creations.

Spray paint might work, drew, or else we could try deploying our own Terminator robot to alter his own past & our future. (Whichever turns out to be the more practical and effective option, I guess.)



Incidentally, these are the top five thread-titles at this moment:

1. anti armed robot tactics...

2. Government Shutdown!!!!

3. Woke up thinking they're going to crash it this week

4. How Bad Is Global Warming?

5. Collapse Culture

No wonder China's leaders are now expressing themselves with unusual explicitness.

and....
not a thread title but


AFP: Powerful Typhoon Wipha heads for Fukushima, Tepco bracing for ‘inflows of water’

oh and let's not forget gold bullets

good points wombat, but I don't think it's too likely they'll ever get 'out of control'

anyway, they should have people on their payroll working to defeat them now, but that's not the kind of sensible thing defense contractors do.

justdrew » Tue Oct 15, 2013 2:51 pm wrote:good points wombat, but I don't think it's too likely they'll ever get 'out of control'

Definitely was not implying that LAR's needed to be "out of control" to represent an existential threat to the human race. They will be used for ethnic cleansing around the world.

In a lot of ways, LAR's going "out of control" is the only good scenario once we're marching down that particular road...

Wombaticus Rex » 15 Oct 2013 12:59 wrote:

justdrew » Tue Oct 15, 2013 2:51 pm wrote:good points wombat, but I don't think it's too likely they'll ever get 'out of control'

Definitely was not implying that LAR's needed to be "out of control" to represent an existential threat to the human race. They will be used for ethnic cleansing around the world.

In a lot of ways, LAR's going "out of control" is the only good scenario once we're marching down that particular road...

one thing worth thinking about, we really could have build such things pretty much any time from say 1970 on, why did it wait so long? (remember they could remote control drone aircraft in the 40s iirc)

Paintball guns.

But surely our (their) soldiers would first need chip implants to id friend or foe, no?

Our only weapon will be the unexpected. AI is bad at that. Humans are good at it, even if unintentionally.

Several things I'm sure anyone has thought for more than a few minutes:

1. War bots can be hacked and turned against their users, and/or each other.

2. War bots are very complicated pieces of machinery that will require maintenance and skilled work to repair and keep running; this makes the fragile.

3. The militaries are already plenty lethal without warbots; the bots seem to only add another dimension to their lethality, but mostly seem like a private industry military industrial complex boondoggle.

4. Many factions will make their own bots and steal IP from each other. I think I saw Syrian soldier controlling a machine gun with a joystick. They could easily get their hands on some software.

5. I fail to see how these things will stand up well to conventional explosives.

6. Offensive weapons cause the creation of defensive weapons, and vice versa. The arms merchants profit constantly churning through first offensive and then defensive weapons. Leonardo Da Vinci did this. He was a weapons designer, amongst other things, and he would sell to anyone that would pay. I imagine people are now developing small arms that neutralize machinery; maybe an EMP gun, or an EMP grenade (Deus Ex, anyone?). Laser weapons, or even hyper accurate weapons like 50 caliber sniper rifles could be used to great effect against bots.

7. With regard to Wombat's concern of 2.0 milbots going on a rampage; if they're made TOO dangerous, than the even a small likelihood of them being subverted would present too great a risk to the power that creates them. So, hopefully an AI omnilethal bot is not in the cards just yet. Humans are lethal enough with the present technology they have.

I wonder if there will be scaenarios of deliberately released 'wild robots' to see what happens, in an analagous way to spraying urban centres with mimics of biological weaponry to observe their impact.
There have already been films on the theme of 'black budget humanoid robot escapes from deep underground research center and hits New York'.

I think that some of the people involved around Transhumanism seem unbalanced enough to sanction the creation of a Terminator style robot... just because they can.

So will Sixth Generation 'Terrorism' consist of Al-CIA-D'uh operatives in caves in Tora Bora, sitting around joysticks controlling Terminator robot fighting G.I. Joes on the streets of New York?

President Ivana Bush-Putinova 2101
"No one could have predicted these American made SafeHomeland robots would start attacking Americans!... with flame throwers! and highjacking Space Shuttles and flying them into DisneyWorld!!!"

seemslikeadream » Tue Oct 15, 2013 1:44 pm wrote:and....
not a thread title but


AFP: Powerful Typhoon Wipha heads for Fukushima, Tepco bracing for ‘inflows of water’

I just finished up work on a project related to DARPA's robotics challenge being held in December, under the false pretense of providing aid and emergency response to a disaster just like Fukushima. However, the participants in the competition include weapons manufacturers like Raytheon and I don't think any participants are ignorant of the fact that their systems will one day be used for war. Companies like Boston Dynamics are at the forefront of DARPA's attention, no matter what the intention or endgame is for war robotics.

My client was prohibited from portraying any kind of conflict, even in a rather playful spirit, by DARPA. There's a very tight control over making the competition appear to be humanitarian, and as cute and cuddly as possible.

In terms of hackability, robotics are still controlled by USB uploaded directly into the machine. One USB contains the information for task A, another for task B, and so on. Seemingly very hackable. I'm not sure that manual "hacks" of optics would work, especially by the time these machines are online and hunting us all down.

sensor confusion may be important. There's no reason why such things should "miss" much, they'll probably be able to hit whatever they shoot at 99% of the time and a human will not be able to zig zag fast enough. (yes, that's an "the in-laws" joke) However, an nice infrared laser should be able to burn out or blind them.

and... decoys. heated articulated mannequins that'll look like people but can take lots of bullets and keep moving.

but eventually humans just don't make the best gun platform. making little remote-by-wire mobile gun platforms might be best, aim-assist software like that new computer-assisted sniper rifle is probably do-able for these units. Might be best to design them as open sphere structures that roll around quite fast. but then we're in fighting fire-with-fire mode.

but I don't know that these things are ever going to be primary. full autonomy is forgetaboutit in less than 20 years if ever, the existing cases only cover sending a device into a 'hot zone' where any target is presumed to be hostile. That's not too tenable really. So most likely these things would assist human soldiers right along side, not run off on their own. and any of these machines will likely remain very vulnerable to AP rounds, etc, pending some significant engine improvements to give them enough power to stack serious armor on. Even tanks remain really very vulnerable.

So in the end, it all comes down to politics, either we get sane reasonable people running the show or we go down in flames, so things remain status quo.

I think once AI starts assisting humans in the programming of other machines, the rate of potentially weaponized robotics development is going to accelerate exponentially.

For instance, this DARPA challenge in December will test robotics teams' abilities to autonomously drive a vehicle, break down a wall, turn a valve, climb over a pile of loose rubble, clear debris, climb a ladder, open a door, and connect a fire hose to a valve and operate it. That these robots are even able to accomplish these tasks surprised even me. From the outside, robotics has been a slow-moving field, but I think they are reaching a point at which their capabilities are going to take a massive leap. AI is a huge unknown in this.