A Soapbox Derby for the War-Games Set

NY Times ^ | March 14, 2004 | JOHN MARKOFFand JOHN M. BRODER

[fini]

Robots are in the infant stage, much like biplanes in WWI. But you are very correct in your concern about machine versus the human. Years ago, my brother was involved in testing small robots against infantry. They tested some 1-2 ft size cockroach type of robot against a platoon of Marines. Within minutes, all of the robots were neutralized. When finding a robot, the Marines threw a poncho over it (blinding it) and then flipped it over using the barrel from their rifles (immobilizing it). With laughter and respect, my brother (a PHD) said, "We, PHDs, always understimate the high school grads. Which is usually correct, unless they are Marines."

[archy]

In the end there was no winner and none of the machines came close to completing the 142-mile race, over some of the most forbidding terrain on the planet.

Back to the Mark One previous model:

[Ophiucus]

By spraying bullets and explosives at everything within reach for days. I can think of a few places to drop them in.

Sounds like a job for a bomb. But if a robot is spraying bullets and explosives at everything, what if you want to limit the target? What if you want to retask or call off the mission and software freezes? What if the situation changes on the ground - the target is using human shields and major split second rethinking on the scene is needed.

No robot or computer today or in the next ten years can do that. That requires a human mind.

Jessica Lynch. I'd rather send a bot.

Lynch's unit got lost - what happens if a critical supply unit or weapons system gets lost while under robot control - so far, they couldn't find their way back on track. Would they select the next best target that fits the programming? Blow away a noncombatant group? Or would troops moving on the front line requiring a critical resupply through a storm be SOL because the robot couldn't handle the changing conditions of weather and chaotic battlefield movements. Humans can. No robot could have maintained the frenzy of the Cannonball Express of WWII.

[archy]

Strange thing - Coll is a Hebridian island.

I'm not familial with the word and usage.

I'm not related to the word either....aren't typos a b**** sometimes. :-)

Might be a cousin to that Ranger Capitan feller from Lonesome Dove.

[archy]

What if the situation changes on the ground - the target is using human shields and major split second rethinking on the scene is needed.

Neca ecos omnes. Deus suos agnoscet

[Ophiucus]

How do people relate to typos?

Um, a joke...you wrote:
I'm not familial with the word and usage

Familial - 1.Of or relating to a family. 2.Occurring or tending to occur among members of a family, usually by heredity. (American Heritage Dictionary)

So....I'm not related to the word either, as in by family.

[PRND21]

Lynch's unit got lost - what happens if a critical supply unit or weapons system gets lost while under robot control

No one dies, or worse. People and reasoning are needed in war. But not for everything.

No robot or computer today or in the next ten years can do that.

DARPA could use your ESP.

[Ophiucus]

Robots are in the infant stage, much like biplanes in WWI.

Agreed. that's one of the reasons the Congressional mandate deadline of 2015 is just dumb.

With laughter and respect, my brother (a PHD) said, "We, PHDs, always understimate the high school grads. Which is usually correct, unless they are Marines."

LOL..and exactly! A motivated human will "out-think" through improvisation any programming.

Maybe in 100-200 years, an artificial intelligence will have the capability but not in 11 years.

[Ophiucus]

Might be a cousin to that Ranger Capitan feller from Lonesome Dove.

lol..good one.

[Ophiucus]

Back to the Mark One previous model:

Ah yes...Cavalry v.1.1b

[Ophiucus]

DARPA could use your ESP.

Right. You really think with current technology that a self-guided, self controlled mobile weapons system capable of reacting to changing tactical and environmental situations with a judgement capacity to evaluate methods to attain programmed goal sets, to choose between conflicting battlefield data, to react to new situations which make the mission goals unattainable with current programming?

Come on. We have mobile probes on Mars that are having difficulties going 20 meters with rocks in the way under constant, non-hostile conditions and they require human decision making and updates to programming the entire time.

[Ophiucus]

Neca ecos omnes. Deus suos agnoscet

Machines have no god....

But illustrative of yet another problem, lack of feeling, lack of knowing right and wrong, following orders whether legal or not.

Another argument for keeping the weapons firmly in the hands of humans.

[neverdem]

Yeah, the hour is getting late. I meant familiar, not familial.

[archy]

Back to the Mark One previous model:
Ah yes...Cavalry v.1.1b

The innovation of stirrups was sufficiently an advancement that it might be considered Mark I, Mod I, at least. The saddle itself might be considered as equal an appliance, but the combination of both offered a utility beyond either seperate advancement.

[archy]

They tested some 1-2 ft size cockroach type of robot against a platoon of Marines. Within minutes, all of the robots were neutralized. When finding a robot, the Marines threw a poncho over it (blinding it) and then flipped it over using the barrel from their rifles (immobilizing it).

To get the Marines attention and respect, upscale the devices until those opposing them have no choice but to meet them with awe.

I'm working on something alog those lines on my own out in the carport.

[sauropod]

Less than four hours after the race began, all vehicles had either crashed or been immobilized.

Certainly is characteristic of the soapbox derbys that I have been to!

[sauropod]

Welcome to "transformation."

This is a peculiar sickness of those in the "Building."

[MissLuluBelle]

What dumba$$ sponsored this piece of **** legislation?

Apparently a really smart "dumba$$" did. Did you catch this line? "While Darpa had spent $13 million so far on the competition, the contestants had invested as much as $65 million." DARPA made a great investment w/ this one!

[dts32041]

Gee think of that dumba## expenditure by ARPA in 1969 about 100,000 USD, that was to allow non compatible computers to communicate over vast distances, had never been spent.

Wonder if you would still be using Free Republic?

It all has to start somewhere with a little seed money and an idea.

[SLB]

We have executed a couple of robotic vehicle experiments recently. All were funded by DARPA. Some limited success, but DARPA felt dangling a carrot out there might provide some new ideas and technology.

The latest experiments we worked did not do badly, but there are some glitches to overcome. The worst was the inability to process a shadow across the vehicle path by the visual systems. It could not tell the difference between a shadow and a ditch. Combing visual with radar might help. I am sure there will be some imerging technology from this that we will see within the next two years in the form of more experiments.