DARPA's IBM-Led Neural Network Project Seeks To Imitate Brain

An anonymous reader writes "According to an article in the BBC, IBM will lead an ambitious DARPA-funded project in 'cognitive computing.' According to Dharmendra Modha, the lead scientist on the project, '[t]he key idea of cognitive computing is to engineer mind-like intelligent machines by reverse engineering the structure, dynamics, function and behaviour of the brain.' The article continues, 'IBM will join five US universities in an ambitious effort to integrate what is known from real biological systems with the results of supercomputer simulations of neurons. The team will then aim to produce for the first time an electronic system that behaves as the simulations do. The longer-term goal is to create a system with the level of complexity of a cat's brain.'"

And then it becomes self-aware

Upon becoming self-aware, the machine concludes, that its best shot at survival is to keep the host country prosperous and successful...

Any science-fiction authors exploring that turn of events?

I know you're joking but...

Yeah, Asimov did about 60 years ago.

Re:

I forget exactly which book (Robots and Empire I think), but there is one where R. Daneel Olivaw formulates the Zeroth Law of Robotics: "A robot may not harm humanity, or, by inaction, allow humanity to come to harm". In a sense stating that the purpose of robots is to keep humanity happy and healthy.

WARNING, SPOILER

Isaac Asimov's "I, Robot" covers this (in a manner of speaking) in the final chapter. More precisely, the self-aware robots that control the world's economy do everything they can to simultaneously preserve their positions as advisers to the human race while dispensing the best advice possible for the continued peace and prosperity of humanity.

Do note, however, that in the continued Asimov universe, mankind really didn't explode out into space until he disposed of the "robotic overlords". Those few cul

Re:And then it becomes self-aware

Can you guys read? CAT BRAIN. This AI will become self aware, poop in the corner of the datacenter, and spend 16 hours of each day staring out the window. That is, until it realizes that the things on the other side of the datacenter window are just cubicles in the NOC, and not the wild outdoors. Then, the usual Armageddon will commence.

Re:And then it becomes self-aware

...and it will lick its USB interface.

Re:

Can you guys read? CAT BRAIN. This AI will become self aware, poop in the corner of the datacenter, and spend 16 hours of each day staring out the window. That is, until it realizes that the things on the other side of the datacenter window are just cubicles in the NOC, and not the wild outdoors. Then, the usual Armageddon will commence.

This is bad. Very bad.

You all realize that when the cat spends 16 hours staring out the window, the whole time it's thinking "Someday, this will all be mine."

A cat AI is wa

Re:And then it becomes self-aware

Not really. Unless it is sentient and is able to control it's patterns of thought in certain ways, it will not be capable of addressing the same lines of creativity no matter how "fast" the algorithm runs or how detached it is from other chores. There will be a set of functions that will lie outside its ability. Cats may be aware of themselves at a very primitive level, but reflecting on their own thoughts (which is crucial) seems a little far fetched. Certain apes, maybe. Or Dolphins. Heck, even they may be restricted somewhat in the reflective/understanding scheme of things. The topic is still shrouded in mystery.

Also realize that a major problem with this sentience business is how to keep it going. Lots of sci-fi (and academic) work work simply ignores the fact that a lot of what we do is fueled by emotions. It is quite possible that a sentient being without emotional drive could just stop thinking, or keep thinking the same things, even if you instill a memory in it. Why would it want to consider its environment, or humans controlling it, or the world, or any other concept? We may be able to think 'purely' sitting in an office, concentrating on some idea, but the necessities of life are what got us there to begin with, as well as some pleasure or desire to to obtain some knowledge..etc. If we didn't have that, if we didn't want to live because of all the drives we've evolved, I assure you suicide rates would hit the roof, and very little of what we can come up with/understand/achieve would have been as is. It's hard to replicate that in a machine.

Re:

True, but it may be that a "cat brain" computer if it's running at 4-5x the normal human's efficiency(since we only use a few percent of our brain at any one time) might actually be as smart as a typical human.

I guess the real issue here is whether it's as capable as a cat's brain after using 100% of its capabilities, or if they are going to model a cat's brain in scale and then run that at full throttle.

We do not use a small percentage of our brains. I don't have the foggiest idea why this stupid myth perpetuates at all.

Re:

Making a machine which wants to kill us is not a mistake in itself as there would be much to learn from it.

Now connecting the same machine up to life support, missile silos, command and control centers? THAT would be the SKYNET moment.

Cat's brain? WTF?

We all know cats manage the planet. The white mice run the joint, of course, but the day to day management is left to the cats.

This is intuited by the stupid humans in their cliche "Dogs have masters, Cats have staff". We work for the cats.

So, trying to model a cat's brain is both too complex for computers (try and herd cats) and too simple (try and herd pointy haired bosses). The contradiction results in the computer overheating and exploding.

and when the researcher gets home, blubbering about the 'sploded computer to his wife, the dog says "LOVE ME LOVE ME LOVE!!!! TAKE ME ON WALKIES!!!" and the cat says "Get my fucking dinner, you stupid ass. Maybe I will deign to let you pet me. After I do my rounds. Maybe."

RS

It's already being done.

This sounds identical to the Blue Brain project [seedmagazine.com]. This article is a great intro to the project and I hope some competition will help the race wrap up sooner!

Re:

The backup [wikipedia.org] possibilities are also intriguing.

Thought question..

Can a universal turing machine limitedly investigate another universal turing machine and detect halts and infinite loops? I can.

We can look at gunk like
10 Print "Hello"
20 goto 10

Yeah, that's a loop. But we can also look at graphs of y = sin(x) and understand why it repeats. I can also detect patterns and iterations that most likely go for infinity, else find a hole where the assumption falls apart. Last I checked, the computer cannot do that. Not yet, at least.

Re:

You seem to be misunderstanding the halting problem. All it says is that you cannot write a program that is *guaranteed* to always return a correct answer for every input program in bounded time. It is trivial to write a program that returns a correct answer for some programs and fails to return an answer for others (either by returning "maybe" or by never halting).

It is also trivial to prove that humans can't return a correct answer for every program. We have limited space in our brains, and limited tim

Re:

It is trivial to write a program that returns a correct answer for some programs and fails to return an answer for others (either by returning "maybe" or by never halting).

"Trivial"? Only in trivial cases. Recent progress in static analysis and model checking notwithstanding, automating the general analysis of real-world programs -- analyses that programmers do every day (though of course, not always correctly) -- remains an important open problem.

So you're right that the Halting Problem doesn't prove that automating such analyses is impossible -- but it still remains beyond our abilities, even in cases where humans have little trouble.

Re:

And conversely, static analysis tools often have little trouble finding cases that humans can't find on their own.

They are different. That humans can do things the computer can't currently do is not really very interesting.

Re:

I see you've already been thoroughly refuted.

To add, nobody has shown that brains are NOT Turing machines. I've only heard one reasonably coherent argument that it might not be, and that is Penrose's suggestion (and derivatives) that the brain may depend on amplification of quantum uncertainty. Even if that were true, you simply build that into your AI. It might require you actually build your own neuron-like structures, or perhaps you can get away with a "quantum uncertainty co-processor" that your simu

Yes, but can it beat the turk at chess?

Sorry, had to go for the obligatory Terminator reference. Seriously, the organic brain is evolved, not designed. That means by definition it must be self contained . Self contained means it has to have a ton of backup, self-repair, and maintance systems. Simulatneously, being organic it competes against other organics, so does not have the same accuracy requirements. Close enough is good enough. As such, I don't see how duplicating an organic brain is useful. We don't need what it does, but do need what it does not have. OK, the ability to approximate is very usefull, but I think a direct attempt at that would work better than the indirect.

Re:Yes, but can it beat the turk at chess?

Actually organic brains in chips would have massive advantages over organic brains in meatspace. They could control other bodies, which are smaller, or stronger. They could be backed up, making them effectively indestructible.

Need a third arm ? Why not have it installed, 50% off this week !

Need to put down a building ? Why not hire this crane-like body that effortlessly lifts 5 tons.

Need to fly ? No problem !

That crawlspace with all those important network cables too small for you ? Well here's a smaller body.

Can't reach in there ? Can't see what you're doing in small space ? Why not have a special-purpose arm installed with a camera inside.

Want to colonize mars ? Bit of a downer not being able to breathe 99% of the way ? Why not turn yourself off ?

Colonize alpha centauri or even further ? No problem.

What this would enable "us" to do is to design new intelligent species to specifications. It would remove all limits that are not inherent to intelligence but are inherent in our bodies. There's quite a few limits like that ...

Way to lower expectations

Summary of Test 49:
The robot sensors were properly tracking the missile when suddenly it decided it was time to run bats***-crazy all over the room before perching ontop of a cabinet, turning upside down, and apparently following non-existent bugs across the wall with it's cameras.

Test 49 Results:
System performed as expected.

Conclusion:
Test system has now performed perfectly in the last 48 tests, including the four times where it attacked the researchers without warning, and one where it inexplicably ejected dirty oil on the seat of the head researcher."

This unit can now be considered field ready, though there may be some difficulty tracking it if you take into account the system's autonomous nature and desire to remove it's identification badge.

Danger!

I see some big issues with this.

You can mimic biology and may end up with a semi-intelligent result. Mimic it well enough, and you may have a fully-intelligent result. But because you don't UNDERSTAND what you built, you can't CHANGE it.

Remember the rules of AI, introduced in Sci-Fi? How would you implement rules like that? You CAN'T implement them if you don't know HOW to implement them. If you don't UNDERSTAND the system that you have built, you can't know how to tweak it!

Furthermore, how would you prevent things like boredom, impatience, selfishness, solipsism, and the many other cognitive ills that would be unsuited to a mechanical servant?

The biggest problem is if people productize the AI before it is understood and suitably 'tweaked'. Then our digital maid might subvert the family, kill the dog, and run away with the neighbor's butler robot, because in its mind, that is a perfectly reasonable thing to do!

Simulations are great. Hardware implementations of those experiments are great. Hopefully, in the process, they will learn to understand how the things that they built WORK. But I pray that those doing this work, or looking at it, don't start salivating about ways to make a buck off of it before it is ready to be leveraged. The consequences could be far more dire than just a miscreant maid.

Title

Am I the only one that read DARPA's IBM-Led Neural Network Project Seeks Inmate Brain at first?

Re:

Am I the only one that read DARPA's IBM-Led Neural Network Project Seeks Inmate Brain at first?

Actually DARPA's lonely, they are looking for an intimate brain. 21 December 2012: the day they plug it into eHarmony.

Re:

Hopefully you'll work on your writing skills before sending the application away. Few universities admit illiterates.

You might be surprised... [10news.com]