DARPA, US Army seek 'social computing' tech

The social computing phenomenon continues to spread, with news now emerging that Pentagon warboffins are commencing work on a military project entitled "Technologies for the Applications of Social Computing (TASC)" - meant to "support leadership decision making at the strategic, operational, and tactical levels". But this is …

[Anonymous Coward]

Get it right, Rozenblit

Although Thermopylae resonates through history, the actual turning points were

a) Salamis

b) Plateae.

Not forgetting the battle of Marathon some years before, during the first attempted invasion of Greece. Well, Athens, actually.

Pedant, moi?

[Michael]

We are a computer simulation

The real people are continuously running trillions of simulations to understand psychology, sociology, consumer behaviour, ethics and everything else. The likelihood of you being one of the true people is so small as to be negligible. Embrace your position as artificial life, and they will be forced to end our false existence.

[Albert Gonzalez]

Psycohistory limits

As anyone whio had read the series, you need at least 1 mol of people to make statistics work ..

This is 6'02 * 10^23 people. As there are now about 6'7 * 10^9 people in the world, you have too little people to have any menaingful statistical answer.

Jou just have to get hold of the population of 10^ 4 planets to work ...

[breakfast]

The fifty-fifty rule

It seems as though half of the things that DARPA are financing are just absolutely crazy, but projects like this are pretty awesome.

Not that I think that what they are trying to do is necessarily possible, but it would certainly be a fun project to work on...

[Tawakalna]

Trantor

you only need one planet as a model for psychohistory, but it needs a population of 40 billions.

[Peyton]

Does this sound to anyone else like?

Some new, fresh-from-college hires have figured out a way to get paid to muck about on Facebook all day?

[Luther Blissett]

Kool-A(c)id [skunk]wurx

> ... the development of new technologies to rapidly create theoretically-informed, data-driven models of complex human, social, cultural, and behavioral dynamics that are instantiated in near-realtime simulations ... technologies of interest include the formalization and semantic representation of social science theories, the semantic integration of disparate types of social science data, techniques for analyzing these data, and efficient computational techniques for rapid data processing.

Sweet talk. I especially love the bit about "formalization... of social science theories". I wonder how many boffins danced on the head of a pin to superstring all that together. Unfortunately they forgot something (apart from reality) - Godel's Proof. In practise, only the last clause is the slightest bit sensible.

[Mike Flex]

Greek snack food

@Bert Ragnarok

> Not forgetting the battle of Marathon some years before

They lost out to Snickers.

[John Smith]

Deathbook/Facekill

Vintage Lewis.

"rapidly create theoretically-informed, data-driven models of complex human, social, cultural, and behavioral dynamics that are instantiated in near-realtime simulations "

For example to construct a model of part (or all) of the Fortune 500 and predict how the traders will behave given their *perception* of how one (or all) shares will move, generating massive profits all round.

Blackadder "So the fact that the answer to this question has eluded the finest minds does not bother you."

Percey "Oh no. I look a challenge."

Blackadder "The lights are on, the windows are open but Mr Brain has left the building."

[Seán]

Aw shit

The last time they did this those Rand idiots almost destroyed civilisation. Just look at the fallouit of cabbageheads clogging up the US consuming without questions.

[Anonymous Coward]

I'm thinking more Philip K Dick here

Let's hone up the model so we can predict who is gonna cause trouble and neutralise 'em n childhood before they become active. A cold clammy feeling at my back, a knockin' at my door....

[Luther Blissett]

RE: you only need one planet as a model for psychohistory

to create a theory, correct. In practise, who can say? Experiments can go wrong. I would say even one planet is one planet too many:- (do you like my dog's bollocks)

1. You only need one psychohistory as a model of this planet. Ouspensky's observation that history is the history of some crime or other, sums it up nicely (if not very nicely).

2. You only need one model as a psychohistory of the planet. Cue irony. Roll up folks, take away your preferred -ISM here - gratis, there's one for everybody, but one per customer only. (Trade-ins accepted). We're all chattering classes now, eligible for bread and circuses while the psychos are getting on with making history. (You don't seriously suppose they give a shit about a/the scientifically correct model?).

3. You only need one model as a planet of psychohistory. God knows! If you're going to have something like Evil/Satan/etc in your grand scheme of things, you only need the one. Any more than one is not greedy, it's just silly. (That's what I call ID).

[Magani]

Waiting for Godot^h^h^h^h^h Hari

Where's Hari Seldon when you need him?

Mine's the one with the Prime Radiant in the pocket

[Huw Leonard]

Fundamental perspectives

I wonder if anyone will bother to let DARPA know that a deterministic, and hence reliable, solution to this problem is impossible. Nah... why kill the golden goose before you've got all the eggs its gonna lay?

[Adrian Esdaile]

Er, hasn't this already happened?

"as a result of my study in markets and economics I applied...

I applied my predictive techniques to the subject of war.

I formulated a US war clock.

The probabilities are 78% that the US will be involved

in a major military confrontation between 1988 and 1992. "

- lyrics from 'Single Gun Theory's "Metaphysical" circa 1994

Mine's the Sony Walkman with Am-stereo and TDK metal tapes!

[amanfromMars]

And where is the UK DARPA Crowd/Cloud Boffinry? Sold off down the river to/in QinetiQ type deals?

.... And/or just a Perfectly Kept Enigmatic Secret working Perfectly ... in IT Cover and Clouds?

"Fundamental perspectives ..... I wonder if anyone will bother to let DARPA know that a deterministic, and hence reliable, solution to this problem is impossible. Nah... why kill the golden goose before you've got all the eggs its gonna lay? " ....... By Huw Leonard Posted Sunday 8th March 2009 07:52 GMT

Huw,

Whilst I might fully agree with you on the present, short sighted, money/honey trapped DARPA methodology/Imaginative Business Model, one would only need to Propose to them/Offer them and/or Register over the Internet, a Special Semantic Discovery and the Building of an Algorithm for Rock Solid State Drive which Deliver Digitally/Virtually the Power of an AIMighty GOD for Total Information Awareness/Perceptions Control and Management, for them to Realise the QuITe Universal Significance in Contributing to ITs Intellectual Property Magic Circle/Collective/Corporative Commune ....... which would be all of those in Combination .... for an AICombine into Greater Great Game Big Picture Productions and ProgramMING Projects ...... as earlier shared in Registered Posts here ....... http://tiny.cc/dzI7i

[4irw4y]

Welcome 2 Hell Jerzy!

See your team @ El Reg also.

So mighty, rich and full of promices those worldwide defence agencies are! Why won't them drop their own rope, build their own datacentres and get the pleasure to arrange their so-called national security inside their own net... and see how many civvies join it? Methinks that the civil web might get into the doggy-style position.

[John Smith]

Needs a better project name

Needs a better project name

"TASC" is a bit bland. How about something a little more definitive.

The "Staff officer in a box" project seems to sum up the sort of data collection / reduction / presentation skills they are seeking to build.

However that might sound a bit threatening to the officer corps.

Mines the one with a copy of Ralph Peters "The War in 2020" in the pocket.