Pentagon To Crowdsource Weapons Software Testing 80
An anonymous reader writes "The Pentagon plans to fork over $32 million to develop 'fun to play' computer games that can refine the way weapons systems are tested to ensure they are free from software errors and security bugs, according to a Defense Department solicitation. The goal is to create puzzles that are "intuitively understandable by ordinary people" and could be solved on laptops, smartphones, tablets and consoles. The games' solutions will be collected into a database and used to improve methods for analyzing software, according to the draft request for proposals put out by the military's venture capital and research arm, the Defense Advanced Research Projects Agency."
Ender's Game? (Score:1)
Mr. Card should be proud.
The Last Starfighter (Score:4, Funny)
If you get really good at the game, they give you a real spaceship to pilot!
Re: (Score:2)
Re: (Score:2)
note when a General says something that might be funny ITS NOT A JOKE (unless you see one of his staffers laugh first)
Re: (Score:3)
note when a General says something that might be funny ITS NOT A JOKE (unless you see one of his staffers laugh first)
"You can't fight in here -- This is the War Room!"
Re: (Score:2)
You have been recruited by the United States of America to defend the frontier against Iran and the Middle East Armada.
Re: (Score:1)
So back on-topic, why would you want to help these assholes? It's amazing how much less military spending you need when you aren't trying to be the world's cops.
you only think its a game (Score:1)
You only think its a game, until you try to log off in mid battle, and some men in black show up at your basement abode and insist that complete the mission. Ender's Game, anyone?
Re: (Score:3)
If you get really good at the game, they give you a real spaceship to pilot!
Indeed and you don't even know about it!
This isn't the first time they've tried this Crowdsourcing thing - there's currently a Abrams tank at the bottom of the middle of the Atlantic, with a great hole blown in its side, which was supposed to be run around Fort Hunter Liggett, in California, based upon use patterns, moderation and metamoderation on /. Which should tell you something ... it'll all end in tears (and give several lucky tube worms a new home.)
Reminds me of... (Score:2)
reading Ender's Game back in grade school.
Re:Does it matter? (Score:5, Funny)
Oh, you meant intact.
Re: (Score:3)
"Help! Help! Mr. Wiggles is stuck in that tree!"
"Don't worry ma'am, we'll get rid of that tree for you." [BLAM] "That will be $90k. Here's your USAF lapel pin and novelty US flag for Mr. Wiggles's funeral and seven cap guns for his 21-gun salute - We'll bill you for those later. You're welcome, have a nice day."
Need Hardware (Score:1)
I'd love to work on the testing software, but I'm going to need some hardware to make sure it interfaces correctly. Just drop off a jet fighter and a couple smart bombs and I'll be set.
Why release it outside the US military, at all? (Score:3)
Why release it outside the US military, at all?
Sure there are some ordinary people in the various branches of the military.
Makes no sense.
Re:Why release it outside the US military, at all? (Score:4, Insightful)
I do weapons safety, including software.
In short, no one in the DoD systems engineering group really gets the big picture here. When you see someone with stars/flag/O-8+ or an SES position touting safety in the DoD, they are always referring to operational safety. Safety as a weapons system design element is typically (70% in USN, 85+% in USMC, Army is even worse) considered red tape, and a waste of money. The exception to this is aeronautics development programs in all branches, with an exception within those groups for UAV's (despite their need for software safety more than anyone). So aero "gets it", except UAV's are still clueless.
Since major contractors swap employees with the DoD regularly, even mature design houses have issues. This is also reinforced by the flow of money. If the DoD PMO doesn't budget for safety in the contract, the contractor isn't going to require it.
Now, remember the above applies to all RDT&E safety. Given what you know of software systems, what do you think will bear the brunt of fielding an immature design: expensive to change HW or cheap & quick to replace SW? How many federal doD program end up short on their budget at the end of the development and integration cycle?
There is the added problem that almost no one does safety research on weapon systems. Pharma, Nuclear, Aerospace, Civil Safety: yes. Weapons... sigh. One of the issues is that if you make explosives in the US you are typically only selling to the federal government. Thus, independent research doesn't happen, either.
Should the information be released? Probably not. But the current state has major issues, and as software is more common in weapons and defense system, the resultant mishap is only going to get worse.
Re: (Score:1)
Sorry for replying to myself, but:
If the system is doing it's job right, it's software development process should be following STANAG 4404 / Joint Software Systems Safety Handbook. They might be handy references if you want to "win the game"
Re: (Score:2)
Because you have to know what you are doing to actively edit software code and make a positive contribution. Those who are having countless amounts of free time are typically low NCO's and enlisted, they have little college education, and have to be ready for their next mission after their buddy's truck was blown up on the last.
This DARPA grant is soliciting for a method of using those idle people to test software without those programming skills. It is not a direct outsourcing of those skills. I don't b
Stealth QA (Score:3)
I doubt that the "games" or "puzzles" would be advertising "This is a DoD test for new software."
Most likely the users will not be aware that it is a test at all.
Re: (Score:3)
If they didn't spell it out, then they would be courting scandal if it ever got out. Lots of people have moral objections to weapons development - to secretly recruit their labor for something they are morally opposed to would be a huge deal.
Re: (Score:2)
> implying that war is a moral practice,
War can be moral. If a leader is slaughtering his people by the thousands, with millions more starving and on the verge of death, is it more moral to sit by and watch this happen, or invade and try to prevent the deaths of millions of innocents? If one state invades another state essentially unprovoked, is it not moral to come to the aid of the state being attacked? To go to war in the defense of another, especially of one who is an innocent, is one of the most moral things one can do.
farmville... (Score:2)
The DoD will soon be partnering with Zygna to test weapons that placate the masses and keep them occupied and off the streets rather than trying to overthrow our friendly dictatorial regimes.
DARPA (Score:3)
DARPA is involved, so read this whole story as: this is a cool idea which we think might have some potential down the line but will probably never happen.
On the other hand, this is a pretty cool idea. Not sure if you can make the puzzles strictly apply to real-life problems and still be fun, though. That in and of itself makes this an interesting idea. If they can overcome that hurdle, there is a lot of potential to this sort of thing.
Re: (Score:2)
It's a very different set of problems, but people have had success getting players to solve protein folding problems for fun with Foldit [wikipedia.org]. Such games take advantage of the enormous human capacity for rapidly finding patterns, compared to computers trying to enumerate all possibilities.
Re: (Score:2)
Wetware still offers impressive performance for some workloads.
Re: (Score:1)
DARPA is involved, so read this whole story as: this is a cool idea which we think might have some potential down the line but will probably never happen.
He says on the internet.
You killed civilians! (Score:5, Funny)
Achievement "Collateral Damage" unlocked!
Re: (Score:2)
In light of the post I saw earlier about "achievements" being added to Visual Studio, Microsoft could make a DoD version and add special military achievements to it. Project types could be:
* Create new weapons system project ...
* Create new flight control system project
Based on the Visual Studio project type, you unlock different achievements when Visual Studio catches you writing bad code.
Acheivement "Take out Nuke Development Capability" (Score:1)
I think they will be most interested in the novel ways in which people leverage different technologies and weapons that are available to them in the game to get one up on their enemies. They could even mock up future tech ideas to see how people use them in the game to determine if they are worth funding. Stats from the game can be leveraged to improve battle doctrine or tactical training scenarios.
WTF? (Score:2)
So let me get this clear: We help them improve the same weapons and other systems they'll be using on us for the NDAA, SOPA, and whatever other unconstitutional laws they have in the works?
This shows how little respect for us they have as well as how stupid they think we really are.
Re: (Score:2)
This shows how little respect for us they have as well as how stupid they think we really are.
I am fairly sure one cannot underestimate the stupidity of the general public.
Think of the average person, now remember that half of everyone is stupider than that.
I'm doing my part! (Score:1)
Are you?
Are you dead yet? (Score:1)
Nope? Ok, wait for the next code release.
Interesting, but DARPA is exaggerating the problem (Score:1)
I take issue with DARPA's assertion in TFA that formal verification cannot be scaled up to work on a modern weapon system. My office has done it for very long time and we are very software-intensive. We and our contractors just had to get smarter as the system became more complex and requirements became steeper.
Nonetheless, I would be interested in the potential of such a process to find sneak circuits and latent problems. Use it during the development process prior to integration and verification.
Re: (Score:2)
".. just had to get smarter.."
code for: We have deluded ourselves that we can handle it by throwing more process at it.
And weapon system are far more complex and require far more rigor then your puny software.
Dali meets Huxley (Score:2)
I don't even know what to begin to think about this. The summary tried hard. So the Gov is going to ...uh... release fun puzzles that analyze how weapons are tested? So what are the licenses on the games themselves? Will they be all locked down by copyright or can we chop them up and do other things with the code and make forks?
Is it all a honeypot?
"Sir, Mr. X. used our game to hack into us."
"Uh, well, we did ask him to look for attack points..."
"Arrest him anyway!"
Stupid. (Score:2)
On one hand, it strikes me as one of those "Let's make something really stupid but plausibly usable, so all our enemies will waste their time trying to duplicate it".
On the other hand, this is just the kind of stupid I would expect from US military.
Re: (Score:2)
That are using millions of people used to improve methods for analyzing software. That's stupid.
And by "That's" I mean "You're".
Re: (Score:2)
lol wut
Stargate Universe? (Score:2)
Looks like someone has watched SGU episode 1.
Old School (Score:2)
"Would you like to play a game?"
Re: (Score:2)
let's play global thermonuclear war
Polybius was a test to see if they can make insane (Score:2)
They wanted to see if they can use photosensitive epilepsy, motion sickness and vertigo as a weapon.
Underway (Score:2)
They wanted to see if they can use photosensitive epilepsy, motion sickness and vertigo as a weapon.
Hence the push for 3D TV, to crowd source testing of said effects on the general population.
Advanced weapon systems will consist of large-screen TV airdrops set to loop Discovery Channel shows on sharks and dinosaurs with large buckets of 3D glasses, said to incapacitate a small town in under 1/2 hour.
Paranoia! (Score:5, Funny)
Anybody play the tabletop RPG Paranoia? The Computer (your friend, your boss, and the head of your government) was always crowdsourcing weapon testing to the player characters.
"Congratulations, citizen. You have been selected to test this box of grenades. To study the optimal grenade design, these grenades have random fuse lengths from zero to ten seconds. Please report your findings with whatever remaining limbs you can."
"Citizen, welcome to the world of high-tech weaponry. The ULTI-3600 assault rifle has a computer targeting system to maximize accuracy. Please note that to prevent friendly fire accidents, the targeting system will verbally ask for no less than five confirmations before taking any shot. To insure that you properly test this rifle it will now be welded to your arm..."
"The new Duo-strike vibro knife is twice as deadly as previous models, because the hilt has been replaced with another vibro blade. Pick it up, citizen. Go on. Don't you want to help The Computer test new weaponry? Or are you a traitor?"
I totally support the way our military is becoming like a dystopian comedy RPG.
Re: (Score:2)
> I totally support the way our military is becoming like a dystopian comedy RPG.
I reluctantly agree, in the sense that at least it's entertaining.
crowdsource? (Score:2)
How many people got a chill when they saw "crowdsource" and "weapons testing" in the same heading? I had missed "software" on first glance. That made it only a little better.
Re: (Score:2)
Gamer?
Re: (Score:2)
Exactly! I'm probably older than you; Kent State is what flashed into my mind.
So that is where Angry Birds came from! (Score:2)
So that is where Angry Birds came from!
In Soviet Russia (Score:1)
They test weapon ON crowd!
Fuzzing? (Score:2)
I can kind of grasp how the amazing parallel computer that is the human brain can solve new problems for something like FoldIt [fold.it], but I can't see how human gamers could improve upon brute force fuzzing and static code analysis of military software. Maybe I have a lack of imagination?
Anyone care to share a vague guess how something like this might work?
what could possibly go wrong (Score:2)
so if everyone who played the game intentionally did everything wrong, or played badly, would they still use the results?
Call me (Score:2)
I will refuse (Score:2)
Good crowdsourcing and Bad crowdsourcing. (Score:1)