Beta

Slashdot: News for Nerds

×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

New "Wet Computer" To Mimic Neurons In the Brain

ScuttleMonkey posted more than 4 years ago | from the not-so-hard-hardware dept.

Science 132

A new type of "wet computer" that mimics the actions of neurons in the brain is slated to be built thanks to a €1.8M EU emerging technologies program. The goal of the project is to explore new computing environments rather than to build a computer that surpasses current performance of conventional computers. "The group's approach hinges on two critical ideas. First, individual 'cells' are surrounded by a wall made up of so-called lipids that spontaneously encapsulate the liquid innards of the cell. Recent work has shown that when two such lipid layers encounter each other as the cells come into contact, a protein can form a passage between them, allowing chemical signaling molecules to pass. Second, the cells' interiors will play host to what is known as a Belousov-Zhabotinsky or B-Z chemical reaction. Simply put, reactions of this type can be initiated by changing the concentration of the element bromine by a certain threshold amount."

cancel ×

132 comments

Eureka! (2, Funny)

Croakus (663556) | more than 4 years ago | (#30729026)

Puts a new spin on computer viruses ...

Re:Eureka! (1)

interkin3tic (1469267) | more than 4 years ago | (#30730486)

Too soon man! These aren't neurons or real cells, and won't be susceptible to viruses (the biological type). Sounds like they're just micelles or maybe liposomes [wikipedia.org] .

Now see, had you waited a few years until they realized that there's no point in reinventing the wheel, that they should just use actual neurons to make a wet computer rather than these things (which presumably don't replicate or repair themselves and probably won't be as efficient), then that would be really funny. Except now it won't be. Ruined it for everyone there.

Bring on Mayor Daley (0)

Anonymous Coward | more than 4 years ago | (#30729038)

The wet mayor needs a wet computer.

Or was that white computer?

If anything comes of this... (3, Insightful)

mcgrew (92797) | more than 4 years ago | (#30729044)

I'll have to change my opinion that we won't ever have true artificial intelligence. A chemical based computer could possibly become intelligent. After all, thought itself is only an electrochemical process.

Re:If anything comes of this... (2, Funny)

spidercoz (947220) | more than 4 years ago | (#30729112)

we need some real intelligence before we can make artificial

Re:If anything comes of this... (1)

interkin3tic (1469267) | more than 4 years ago | (#30730362)

we need some real intelligence before we can make artificial

So I take it you just heard Sarah Palin is going to be on Fox news, huh?

Re:If anything comes of this... (2, Funny)

phantomcircuit (938963) | more than 4 years ago | (#30729118)

Is it really a computer still though?

Re:If anything comes of this... (2, Informative)

keatonguy (1001680) | more than 4 years ago | (#30732704)

But of course! A computer is not defined as an array of silicon transistors, a computer is any device which processes information. That includes everything from mechanical calculators to the human brain.

Re:If anything comes of this... (5, Insightful)

pete-classic (75983) | more than 4 years ago | (#30729252)

Wow, talk about carbon bias!

Can you cite a reason why silicon-based systems shouldn't be as capable carbon-based ones? Silicon-based have developed at a blistering pace as compared to the carbon. (Though I admit that they have the advantage of actually having intelligent designers . . .) I mean, life has a head start of a few billions of years!

-Peter

Re:If anything comes of this... (4, Insightful)

Alzdran (950149) | more than 4 years ago | (#30729470)

I can't speak for the GP, but I think we'll exploit properties we don't fully understand (say, by growing neurons on a grid that interfaces with them) much faster than we'll be able to translate those properties into other systems.

Re:If anything comes of this... (2, Informative)

Anonymous Coward | more than 4 years ago | (#30730014)

You got it. This is my position: "AI" will first come from neurons that are attached to computer, and able to use that computer as part of its "mind".

Re:If anything comes of this... (1)

icebike (68054) | more than 4 years ago | (#30731154)

I admit to a Carbon Bias.

Wanna make something of it Mr Sandman?

Re:If anything comes of this... (1)

smidget2k4 (847334) | more than 4 years ago | (#30731238)

Silicon based systems are still terribly sequencial. While they do nicely augment what our brains are bad at (sequencial problems), they are absolutely terrible at what ours brains can do amazingly easily (Computer Vision, Strong AI, decision making, pattern recognition, face detection, path finding, etc).

Not to mention our brain balancing and controlling tens of thousands of bodily functions all at once and still having plenty of time to do other things. On the flip side, we know how silicon works, but we have no frackin' clue how the brain actually works.

Re:If anything comes of this... (2, Insightful)

pete-classic (75983) | more than 4 years ago | (#30731596)

Again billions of years head start! If something like human intelligence arises from silicon in fourty million years (the clock started, what? Sixty years ago?) then it will have happened in only one percent of the time as carbon.

I think that you grossly underestimate our understanding of chemical building blocks of cognition. But, putting that aside, I think your argument recommends our efforts on the silicon front. Again, in only a few short decades we have gone from purely sequential (serial) designs to very fast serial designs that can usefully mimic parallel designs, to today's shaky steps toward significant parallelism.

-Peter

Because the evidence. (1)

weston (16146) | more than 4 years ago | (#30731800)

Wow, talk about carbon bias!

Can you cite a reason why silicon-based systems shouldn't be as capable carbon-based ones?

Because there's no evidence thus far for consciousness and cognition in anything other than carbon-based wetware.

You can hypothesize that consciousness and cognition are just another kind of computation, and a lot of people do, which near as I can tell is how we get the idea that silicon-based systems will someday do cognition and maybe even self-awareness when we find the right algorithm. But there are no such systems at the moment, and there's no particular evidence that given hypothesis is correct. It may well be self-aware intelligence is tied to the particular mix of phenomena that take place inside of carbon brains.

Re:Because the evidence. (2, Insightful)

pete-classic (75983) | more than 4 years ago | (#30733890)

Because there's no evidence thus far for consciousness and cognition in anything other than carbon-based wetware.

I'll go further and stipulate that such a thing doesn't exist. But that in no way answers my question.

But there are no such systems at the moment, and there's no particular evidence that given hypothesis is correct. It may well be self-aware intelligence is tied to the particular mix of phenomena that take place inside of carbon brains.

Sure. And we can "maybe" away any idea that hasn't yet been proven in practice. But that doesn't seem like a useful pursuit to me.

But we have been steadily progressing at developing more complex and capable silicon/electronic systems for the last few decades, and there's no particular reason to expect that progress to slow or stop.

On the other hand, we've been working on Biology and Chemistry in earnest for a couple of hundred years, and the capabilities of those disciplines doesn't seem to be generally developing toward any sort of artificial intelligence. We do know for a fact that biochem can give rise to intelligence, but we seem much further from creating it artificially with those tools.

Your argument forcibly brings to mind those who insisted that, while heavier-than-air flight was possible for birds, it was utterly unachievable for man. Their only evidence being that man hadn't yet achieved it. I freely admit that it may not be possible for intelligence to arise from non-carbon systems, I'm merely pointing out that we've hardly begun the attempt.

I doubt that either of us will be proven correct in our lifetimes, but that in no way negates my argument.

-Peter

Re:If anything comes of this... (4, Funny)

plopez (54068) | more than 4 years ago | (#30729296)

Then you'll get computers with chemical imbalances. In other words suffering from depression. It gives the phrase "the computer is down" a whole new meaning.

As a preemptive strike:
http://en.wikipedia.org/wiki/Marvin_the_Paranoid_Android [wikipedia.org]

Re:If anything comes of this... (1)

MichaelSmith (789609) | more than 4 years ago | (#30729352)

No that was because of the terrible pain in all the diodes down his right side.

Re:If anything comes of this... (1)

WCLPeter (202497) | more than 4 years ago | (#30731334)

It gives the phrase "the computer is down" a whole new meaning.

Let's hope that's the least of our worries, I don't even want to imagine what a BSOD error would look like.

Re:If anything comes of this... (4, Insightful)

icebike (68054) | more than 4 years ago | (#30729338)

I fail to see why you need chemical based computers in order to construct artificial intelligence.

One could build the system out of tinker toys and achieve the same results, at different speeds and different costs.

There is nothing that signifies intelligence which is provided by one construction method that is not present in another. Electrical, Optical, Mechanical, Chemical, Pneumatic... They are all just a means to an end.

Re:If anything comes of this... (3, Insightful)

Monkeedude1212 (1560403) | more than 4 years ago | (#30729566)

Chemical reactions have a sort of random-ness to them that electricity through a wire can't duplicate. When the circuit isn't complete, electrons aren't moving. When two chemicals aren't reacting, their molecules still shift about in either their gaseous or liquid form. They could be affected by anything that comes into contact with them, depending on the substance it could be magnetic and thus making their movement affected by all sorts of things.

Think of the number of random events that can occur on the cellular level.

Computer software has gained the ability to learn, and gained the ability to change itself, even learned how to reproduce itself.

What it hasn't gained is the ability for abstract thought, which I attribute to the incredible amount of random events that go on inside our brain.

Though I could just be blowing smoke, I'm no physicist or chemist or biologist.

Re:If anything comes of this... (3, Insightful)

sznupi (719324) | more than 4 years ago | (#30729692)

That mostly signifies how relatively efficient are processes on which our bodies rely, in regards to density of information and computation. Much more than our current silicone-based systems.

But both are far from possible maximum theoretical densities, efficiency. And you have a lot of "randomness" deep down...

Re:If anything comes of this... (1)

Chris Burke (6130) | more than 4 years ago | (#30730118)

Chemical reactions have a sort of random-ness to them that electricity through a wire can't duplicate.

Why can't "a wire" duplicate them? A single wire in a computer doesn't, but that's by design, not fundamental nature.

If the problem with our algorithms is that they're too deterministic, why can't we introduce randomness? Some exceedingly successful algorithms already do.

I surely believe that our current algorithms are not sufficient to produce the kind of "intelligence" we're talking about. However I can't see any compelling reason why it should be impossible.

Re:If anything comes of this... (2, Interesting)

Monkeedude1212 (1560403) | more than 4 years ago | (#30730620)

But we don't actually have true randomness (though you could argue neither does nature, perhaps quantum mechanics is all a bunch of hoopla and the very outcome of the universe can be determined, but we'll save that for another day).

Point is, a random number generator when rebooted will generate the same random numbers when rebooted in the same scenario. Tested myself across many lanuages - I have yet to see a true random number generator.

Our algorithms are VERY deterministic. Lets say there is a purple car, but in my memory I remember it as blue. A slight difference in colour. As computers determine colours in that RBG kind of Fashion, we'll say my memory added a 50 to the red factor, just to put it in understandable terms. Why did this happen? When did this happen? Surely 2 moments after seeing the car I would remember it as blue, but later that week I recall it differently. Was it in the process of putting it into long term memory, or did it degrade over time in long term memory? But why is it that I can correctly remember the colour of my Grandfather's eyes, even though he is long gone?

Lets say for the sake of arguement you put it in your algorithm to on occaison 'randomly' alter data when its transfered to long term memory, and/or when it sits in long term memory. How often do you execute this? Is that up to 'random' chance? And how much gets altered? Is that random too? Could not my entire memory become distorted? Would that be the same thing as a mental disease - the computer happens to get a bad long series of random?

It all comes down to properly producing random events - something which computers don't really do, its pseudo-random [wikipedia.org]

The way to achieve true randomness, (From the article): One measures some physical phenomenon that is expected to be random and then compensates for possible biases in the measurement process.

Or in other words, cellular interactions, chemical reactions, even rolling a die would be better than building a computer to "generate" randomness.

Re:If anything comes of this... (1)

icebike (68054) | more than 4 years ago | (#30730810)

You needn't randomly alter the data, you need only randomly desensitize the selection filter.

Chances are if you were hypnotized you might actually remember the car in its true color of purple, and also the blue dress of the attractive driver.

Desensitizing the filters (blue and or purple and or car and or dress) and use that to hit the database any you have perhaps something similar to what is going on in the brain.

As for the randomness of random number generators it matters not a wit what they cough out when rebooted. It matters only what they cough out NEXT. There are many good scenarios for getting more randomness out of random number generators, and the literature is full of such things.

However, there is no strong evidence to believe that what happens in the brain when associating inputs with memories is in any way "random". If it was you would be as likely to remember a slice of bread when you thought of the car.

So all of this discussion of random number generators is "up the wrong tree barking", a quibble over an unimportant tangent.

So, computer + lava lamp = groovy intelligence (1)

Chris Burke (6130) | more than 4 years ago | (#30731234)

But we don't actually have true randomness

Sure we do, if you actually care and want it. All you need is a source of true entropy.

I have yet to see a true random number generator.

BAM! [hackaday.com] And as of clicking that link, you can no longer say that.

Note that the actual implementation doesn't use a lava lamp anymore. For a general discussion of sources of true randomness, see here [wikipedia.org] . The number generated from these devices are as random as the physical processes behind them.

Our algorithms are VERY deterministic. Lets say there is a purple car, but in my memory I remember it as blue.

Not always. Genetic algorithms (and quite a few other heuristic algorithms) are based on randomness. Even a simple neural network with feedback will exhibit the behavior of "remembered" values changing over time. You can easily combine the two techniques as well to create a very potent "fuzzy logic" system that's only as deterministic as your random number stream, minus that you can be pretty sure that it will converge on a solution to your problem (much like natural selection driven by random changes converges on solutions to the problem of survival).

Lets say for the sake of arguement you put it in your algorithm to on occaison 'randomly' alter data when its transfered to long term memory, and/or when it sits in long term memory. How often do you execute this? Is that up to 'random' chance? And how much gets altered? Is that random too? Could not my entire memory become distorted? Would that be the same thing as a mental disease - the computer happens to get a bad long series of random?

Well that depends on the algorithm, now doesn't it? If it's some kind of genetic algorithm, then randomness is introduced in things like the mutation rate. In real life, mutations are random, but the mutation rate itself is not a random variable (though it does change depending on environment).

If we're talking about something like brute-force simulation of the brain, then the randomness would only be introduced at those points where it must in physics, like qm waveform collapses.

If your point is that we don't know the proper algorithms yet, then I'm in total agreement. But we're talking about why a computer can't be intelligent, not why we're unlikely to make one that is any time soon.

The way to achieve true randomness, (From the article): One measures some physical phenomenon that is expected to be random and then compensates for possible biases in the measurement process.

So you knew about these techniques, but for some reason think they're cheating or something? If the computer has a sensor reading thermal noise and it uses that to produce Intelligence, suddenly it doesn't count as the computer being intelligence? Like what, the little sensor or the hot piece of metal is intelligent but the computer isn't? Come on.

Okay, assume the computer has a source of true randomness. NOW why can't it be intelligent but chemicals can?

Re:If anything comes of this... (1)

vivian (156520) | more than 4 years ago | (#30732598)

Actually electronic systems can indeed have randomness - it's called noise.
There's a lot of effort put into designing existing logic gates etc in computers to ensure that they are not switched off/on from noise, and this gets harder and harder as the designs get smaller - because our current designs depend on say, A & B always generating exactly the same result. This also means that the logic circuits have to use higher voltages, use more power to get above the noise threshold.

A different approach would be to have multiple gates all doing the same computation, but working much more closely to the noise threshold, and thus using much less power - each individual gate might be much less reliable (say, 70% accurate), but if you averaged hundreds of them, the accuracy can be improved. This is exactly the approach that is being taken by some dude from africa who has been trying to create new neural network chips that use way less power, with much more speed, than a conventional CPU trying to implement a neural network in software. More info here: here [discovermagazine.com]

Re:If anything comes of this... (1)

electrons_are_brave (1344423) | more than 4 years ago | (#30732750)

Lets say there is a purple car, but in my memory I remember it as blue. .. Why did this happen? When did this happen? Surely 2 moments after seeing the car I would remember it as blue, but later that week I recall it differently. Was it in the process of putting it into long term memory, or did it degrade over time in long term memory?

But it would never be random. You are making a wrong assumption about how human memory works: that everything we see is transferred into long term memory, although sometimes that information is degraded or retrieved wrongly. But only a percentage of what happens to us is remembered. It islikely that you don't remember the colour of the car you saw at all (wrongly or rightly). All you remember is that you saw a car, maybe of a particular type. You then randomly retrive an image of that type of car from a memory of a completly different time. That car was purple.

So the computer would need to forget some percentage of what it saw but be driven to fill in the blanks of the missing details from other, similar memories. Plus, it would need rules about how certain categories of what it saw would be more or less likely to be forgotten (i.e. it's common to forget cars we see, it would be surprising if you forgot an elephant strolling down your road). Plus it would have to have some things it remembers being very easy to retrieve (the elephant), some things hard to retrieve (the name of that man you worked with last year). Plus you'd need to make sure the computer made a certain percentage of retrieval mistakes (You say his name was Bill but it was Kevin).

Re:If anything comes of this... (1)

mad_minstrel (943049) | more than 4 years ago | (#30730218)

I thought we have invented random number generators already?

Re:If anything comes of this... (1)

Monkeedude1212 (1560403) | more than 4 years ago | (#30730410)

Even those are not truly random.

Load up VB. Simple form, tell it to print a different random number 100 times. Take note of the list. Good spread.

Stop the program. Tell it to run again. List will be identical.

Re:If anything comes of this... (1)

skylerweaver (997332) | more than 4 years ago | (#30730564)

While pseudorandom number generators are deterministic (LFSR, etc), there is no reason that we can not implement a true random generator in hardware.

There are many ways to do this. For example, amplify the noise from a resistor, quantize it, and use the LSBs. These should be random.

Re:If anything comes of this... (1)

Monkeedude1212 (1560403) | more than 4 years ago | (#30730878)

Yes - and to generate thousands of those each instant and compute them would be taxing on the system - and expensive. Whereas this project is attempting to build both the random generator AND the computational hardware required out of an existing model.

Because by the definition of Random-ness in nature, I mean that everything is affected by their surroundings. I could have a different mood based on my temperature. Truly you could program an algorithm for that, but what about Barometric pressure? Okay now how about my breakfast this morning. Or dinner last night actually! Or how about my childhood. Or how about plans for the future? How about how my boss was talking to me this morning? My DNA?

Point is - the amount of random events that affect us is absolutely staggering. But we know they affect us and its not just a simple matter of "random chance". I could have heireditary depression, or I could not. It's a random chance, to be sure, but it is highly influenced by my ancestors. And that might affect my mood severely.

If you still think that getting a computer to not only run a simulation but also generate all the random events that effect that simulation is still a plausible answer with computer technology today - I don't know how you arrived at that answer.

This wet computer is an attempt to properly duplicate only but a mere fraction of what actually randomly affects a situation. If we can get the random parts built-in into the hardware and not determined by the software, then we are one step closer to True AI. After all, the world in which the wet computer would reside would be just as random as our world, and a good chunk of how it would think has already been written and tested.

Re:If anything comes of this... (1)

skylerweaver (997332) | more than 4 years ago | (#30731298)

If you still think that getting a computer to not only run a simulation but also generate all the random events that effect that simulation is still a plausible answer with computer technology today - I don't know how you arrived at that answer.

I merely only implied that it is possible to generate random numbers, which is very useful.

Perhaps you want to simulate something very specific and you know that the random part of the simulation has very specific characteristics (e.g. transient noise in SPICE), this is something that is easy to implement in software.

Re:If anything comes of this... (1)

icebike (68054) | more than 4 years ago | (#30730854)

Visual Basic? Are you for real?

Don't you seed your random number generator with something like the current millisecond or something?

Re:If anything comes of this... (0)

Anonymous Coward | more than 4 years ago | (#30730940)

Shhhhh, he's having FUN WITH COMPUTERS! If you point out something wrong, you'll hurt his enthusiasm, which is pretty much all that's keeping him afloat these days. Well, his enthusiasm will be hurt until PUDDING TIME! which will cause it to go right back up there until nappy nap.

Re:If anything comes of this... (1)

haderytn (1232484) | more than 4 years ago | (#30730350)

Lots and lots of smoke.

Re:If anything comes of this... (1)

Cyberax (705495) | more than 4 years ago | (#30732120)

"Chemical reactions have a sort of random-ness to them that electricity through a wire can't duplicate. When the circuit isn't complete, electrons aren't moving."

Two words: thermal noise.

Re:If anything comes of this... (1)

LockeOnLogic (723968) | more than 4 years ago | (#30730092)

The thing is, although you CAN make a computer out of tinker toys, it isn't the most efficient method of accomplishing it. Given that a brain is far more complex than anything humankind has ever produced, it seems presumptuous to assume that our current methods of computation are ideal for the creation of AI. The intention here is to study the relative merits of mimicking neurons as a method of computation vs. our current transistor based designs.

Re:If anything comes of this... (1)

oljanx (1318801) | more than 4 years ago | (#30730606)

I'm just speculating here, but I'd guess that it would be easier to create a massively parallel processing computer using this approach. Eventually at least. Right now the best we can do is simulate neurons on what are largely linear processing systems, which isn't very efficient.

Re:If anything comes of this... (0)

Anonymous Coward | more than 4 years ago | (#30732444)

look..nobody's judging you terminator. WE know ur a cyborg sent back from the future and u want to be treated equally and all,
nut .. wet jelly-like protein things are flexible, maliable, dynamic things that can grow and learn.

Re:If anything comes of this... (3, Insightful)

MichaelSmith (789609) | more than 4 years ago | (#30729378)

thought itself is only an electrochemical process.

Thats what we think...

Re:If anything comes of this... (1)

Hurricane78 (562437) | more than 4 years ago | (#30733038)

And as a bonus objective, try to prove that anything, except for yourself, exists at all. ^^

Re:If anything comes of this... (1)

Prune (557140) | more than 4 years ago | (#30729386)

Your comment doesn't make sense. Even neuroscientists (as opposed to computer scientists) will tell you that the type of process, electrochemical vs any other, has no impact on the ultimate information processing ability, which can be abstracted from the process which is but mere mode of implementation. A sufficiently complex biological or chemical computer will be Turing-complete, just like an electrical computer, and _can be no more_ powerful--just more or less efficient at various tasks in terms of time or energy or whatever. No one working with neurons has any illusion that the information processing of the neurons indelibly depends on electrochemical actions; the same information processing can be replicated by any other process such as mechanical, electronic, or whatnot. Saying that "thought is only an electrochemical process" is doubly misleading because 1) it implies that the complexity of though have to do with electrochemical reactions when the correct level of abstractions is information processing, and 2) because it implies that only electrochemical processes can represent thought.

Re:If anything comes of this... (3, Insightful)

Chris Burke (6130) | more than 4 years ago | (#30729520)

A chemical based computer could possibly become intelligent. After all, thought itself is only an electrochemical process.

If you believe that thought/sentience/intelligence is only an electrochemical process, then why do you believe that the same effect can't be achieved from a purely electrical process?

Imagine a computer no different than those we have today, but arbitrarily more powerful. On this computer is running a perfect simulation of the human brain, down to modeling every quark. Every process, chemical, electrical, and otherwise that takes place in the brain takes place in the simulation. We can already do this for small numbers of molecules; the only thing missing is the processing power. Why exactly could this simulation not exhibit the same intelligence as you or I? What is it missing, and why can what is missing not be added to the model?

Re:If anything comes of this... (2, Insightful)

sznupi (719324) | more than 4 years ago | (#30729774)

Soul, naturally ;)

Re:If anything comes of this... (2, Insightful)

Chris Burke (6130) | more than 4 years ago | (#30730010)

Soul, naturally ;)

James Brownatron, Godfatherbot of Soul, disagrees! He stays on the scene like a sex machine with 99.99999999% uptime!

I doubt however that's what someone saying "thought is just an electrochemical process" was going for. I mean, I even believe in souls, but I think they're a cop-out for explaining how brains produce intelligence. The question is what does produce intelligence, and I have a hard time believing it's the constituent components, and not the emergent pattern. The issue with AI is one of algorithms, not something that only chemicals can do and computers can't.

Re:If anything comes of this... (1)

sznupi (719324) | more than 4 years ago | (#30730172)

...I have a hard time believing it's the constituent components, and not the emergent pattern.

It's probably somewhere in between (if understanding "constituent components" as relatively simple blocks giving particular functionality, each "generation" building on simpler blocks). Hypotheses relying primarily on emergent patterns...well, I'll put it his way: I absolutely love holographic theory of brain. Adore it. It's mindbogglingly beautiful.

And...our current evidence suggests it's most likely incorrect. :(

Re:If anything comes of this... (2, Insightful)

Chris Burke (6130) | more than 4 years ago | (#30730286)

It's probably somewhere in between (if understanding "constituent components" as relatively simple blocks giving particular functionality, each "generation" building on simpler blocks).

Take it more to mean the functionality itself, as opposed to the molecules that provide that functionality. Basically, however neurons etc work together to make a working brain, why is it so essential that it be physical potassium ions and physical neurotransmitters to create that functionality, rather than electronic equivalents? If you simulate the flow of ions, what's the difference? That's the basic question I'm asking here.

And...our current evidence suggests it's most likely incorrect. :(

I don't really understand... How can intelligence not be an emergent property? A neuron, though fairly complex, isn't intelligent. 2 tied together isn't either...

Re:If anything comes of this... (0)

Anonymous Coward | more than 4 years ago | (#30732460)

What's missing is positive and negative feedback loops. For AI to be emergent, there must be a number of different subsystems communicating, which allows the overall system to "learn". Kickstart an evolutionary process on silicon-based lifeforms and intelligence and other elements will arise.

Re:If anything comes of this... (1)

BlueBoxSW.com (745855) | more than 4 years ago | (#30729624)

Building an AI program based on neurons is like writing a book review by counting describing the length and frequency of words in the book.

Of the flip side, someone got a sweet grant and job security for them and their peers. Whoo Hoo!

Re:If anything comes of this... (1)

Antiocheian (859870) | more than 4 years ago | (#30729818)

You mean that an electrochemical Kismet is upon us ?

Re:If anything comes of this... (0)

Anonymous Coward | more than 4 years ago | (#30730082)

After all, thought itself is only an electrochemical process.

Really? Presumably you have proof of this?

Re:If anything comes of this... (2, Insightful)

phantomfive (622387) | more than 4 years ago | (#30730306)

Biological brain processes are so agonizingly slow, though. It's not just the electrochemical signal, it's the process of learning new things, which involves not only making an electric link, but actually growing new physical connections, extending dendrites and growing synapses. This takes time, energy, and nutrients for the body. Doing it in electricity is so much easier, although we pay for it in energy costs. Compared to a 150 watt computer power supply, the average human body burns around 2000 calories in a day, that's about 2.3 watt hours. Our brains use significantly less energy, but they are so slow!

I am not against research, I am in favor of studying everything in the world, but I don't think it is reasonable to assume that only a chemical based computer could produce artificial intelligence. Chemical-based has significant disadvantages when compared to silicon.

Re:If anything comes of this... (1)

babblefrog (1013127) | more than 4 years ago | (#30731636)

You used small calories instead of kilocalories. That should be 2300 watt-hours. Or about 100 watts.

Re:If anything comes of this... (1)

phantomfive (622387) | more than 4 years ago | (#30731776)

Oh, thanks for the info

Re:If anything comes of this... (2, Funny)

Nefarious Wheel (628136) | more than 4 years ago | (#30730318)

Well, there is precedent. You see, when a mommy computer and a daddy computer love each other very much, they show this sometimes with a special dance. And if they are compatible, sometimes they will plug their interfaces together and engage in in a high bandwidth conversation. And much, much later, out of Mommy's USB port pops a very special process controller, just like you!

Re:If anything comes of this... (1)

AC-x (735297) | more than 4 years ago | (#30731454)

only an electrochemical process? Maybe not... can't find a link to the article but read an interesting piece a while ago on certain structures in nerve cells that trap electrons and seem to behave like quantum computers, so it may be the case that nature is already tapping into quantum computing for thought and consciousness, imagine the brain as billions of networked quantum computers - it's no wonder AI hasn't caught up yet.

Re:If anything comes of this... (4, Insightful)

LUH 3418 (1429407) | more than 4 years ago | (#30731526)

Many people bring forward this idea. I think it stems from the fact that "traditional" AI (which has only really been around for 60 years or so) has not yet yielded a "sentient" computer. People feel that this somehow means traditional AI, and even our whole computational model can't yield sentience. They attribute intelligence to the fabric rather than the logic it implements. I think these people fail to realize that whatever computation biological brains implement, we could simulate it on traditional computers, *if we even knew what is being computed*. The problem is that, so far, beyond the first layers of our visual system, and some very simple systems, we know not much about the way the brain is connected. However, from what's been discovered in neuroscience, it seems pretty clear that the early layers of the visual cortex perform simple convolutional operation that do not involve quantum physics, or fancy shmancy things we couldn't do *more efficiently* with silicon.

The human brain is very complex, but given enough time, we very well might get to understand what makes us sentient and be able to replicate it in a computer. My personal opinion is that the brain is full of specialized hardware that has evolved over a very long time, and helps us to specific tasks (eg: facial recognition, hand-eye coordination, obstacle avoidance, language decoding), with a very powerful abstraction logic built on top (the stuff that "makes us sentient"). This abstraction logic is possibly very complex, and perhaps too difficult for us to conceive of at this time. If we are to learn anything from the rest of the brain, most of this logic probably focuses on transforming perceptual information into a form that makes it easy to reason with. On top of this, we probably again have specialized mechanisms, to do things like deduce causal relationships and generate hypotheses or semi-random associations of concepts (creativity).

The reason the "traditional AI" camp hasn't succeeded at making sentient machines are multiple, but I would sum them up as follows:
1) They have mostly given up. You probably can't get funding for claiming you'll come up with HAL9000, you'll sound like a wacko. Current AI research focuses simple learning problems (i.e.: supervised learning, reinforcement learning).
2) The approaches tried in the past focused purely on formal logic, which, as we now know, works badly in open-ended environments. For it to work well, the properties of the environment have to be simple, restricted and well-defined.
3) Supervised learning, unsupervised learning, etc., will not yield sentience. These approaches, which may actually exist in the brain, are good at solving problems of limited scope only. Our brains are not big wads of neurons performing a single computation. They are much more intricate and integrate many specialized components.

The "right" approach to AI is probably an overall approach, integrating many existing techniques into one system. Perhaps an "engineering" approach to AI would work better. Focus on constructing it and then refining it, as opposed to developing an overall theory of how it will work first and trying to reduce it to its simplest component. We already have computer systems that do speech synthesis, speech recognition, facial recognition, depth perception, 3D model reconstruction, etc. We also have unsupervised learning, supervised learning, reinforcement learning, fuzzy logic, knowledge bases, automatic theorem provers, etc. It should be possible to build a non-completely stupid AI, if one combined all these techniques in the appropriate way. How to connect them, however, is probably where the true AI problem resides.

I for one... (1)

Trecares (416205) | more than 4 years ago | (#30729066)

welcome our new wet computing overlords.

Re:I for one... (1)

mcgrew (92797) | more than 4 years ago | (#30729096)

My daughter's cat made a wet computer. Unfortunately the cat urine shorted the thing out...

Based on bromine (1)

TrashGod (752833) | more than 4 years ago | (#30729102)

Based on bromine [wikipedia.org] , a dark, red, fuming, toxic liquid with a choking, irritating smell; from the Greek meaning stench. I'll take two, please.

Re:Based on bromine (1)

MonsterTrimble (1205334) | more than 4 years ago | (#30729926)

a dark, red, fuming, toxic liquid with a choking, irritating smell

Has anyone seen Keith Richards lately? They might have stolen his brain...

Then again, would anyone notice?

1.8M Euro? (1)

MSBob (307239) | more than 4 years ago | (#30729158)

That's a joke for funding. A project as ambitious as this cannot get much accomplished with a couple of million eurobucks. Ten times that amount would have been respectable. 1.8 million is money than it takes to open a fast food franchise joint in some cities.

Re:1.8M Euro? (3, Insightful)

blueg3 (192743) | more than 4 years ago | (#30729482)

600k Euros / year isn't bad for an exploratory research program. The article suggests that they're funding multiple such efforts, searching for promising ideas that can be furthered by later funding -- which is often how these things are done.

Re:1.8M Euro? (1)

icebike (68054) | more than 4 years ago | (#30729500)

That's what I was thinking too.

If they are trying to make a replacement for a single transistor, fine. But aiming for a whole computer made of chemical components for 1.8M euros seems ridiculous.

Sounds like just enough money to be pissed away in a year, and milked for another grant with "amazing progress" reports just before the funds run out.

Don't see too many mentions of B-Z on /. nowadays. (-1, Offtopic)

Anonymous Coward | more than 4 years ago | (#30729176)

Just thought I'd flag it... http://bzflag.org/ [bzflag.org]

I'll be here all week, try the veal, etc. etc.

great, boffins (2, Funny)

marcuz (752480) | more than 4 years ago | (#30729232)

now just take a beowulf cluster of these and ... what would you do?

Re:great, boffins (0)

Anonymous Coward | more than 4 years ago | (#30729408)

can they run linux?

Re:great, boffins (-1, Flamebait)

Anonymous Coward | more than 4 years ago | (#30729836)

If they're smart, they'll run FreeBSD instead.

Now watch this get modded flamebait.

Re:great, boffins (0)

Anonymous Coward | more than 4 years ago | (#30730114)

slay a beast?

Re:great, boffins (1)

ArsonSmith (13997) | more than 4 years ago | (#30730292)

???
Profit

naked and petrified.

what you say?
take off every zig!

goatse

frist post!!

Re:great, boffins (1)

gstoddart (321705) | more than 4 years ago | (#30730746)

now just take a beowulf cluster of these and ... what would you do?

It's called Slashdot. :-P

Cheers

Research Implications: (1, Funny)

Anonymous Coward | more than 4 years ago | (#30729328)

Heuristics is expected to make a big come back in Artificial Intelligence research, as the machine represents everything in terms of bromides [etymonline.com] .

Silicon, carbon. Feh. Bits is bits, guys. (1)

gestalt_n_pepper (991155) | more than 4 years ago | (#30729436)

I just don't see how this is going to be significantly different than something done in silica. It's the organization and processing, not whether it's made of coal or sand.

Re:Silicon, carbon. Feh. Bits is bits, guys. (1)

dj_tla (1048764) | more than 4 years ago | (#30730418)

Bits is not bits in the brain. We don't have a bunch of neurons representing ones and zeros.

A neuron is either spiking or not spiking, but everything is very temporal; with some stimulus (a particular input current) a neuron might spike at a rate of 100 Hz, with another input, at 20 Hz. How do we translate these spikes into information we can use? Well, we're trying to figure that out.

In addition, in the brain, we do not have the cleanness of digital processing. There is lots of noise from a variety of sources, so a lot of effort is taken to be resistant to noise.

Does it matter if things are carbon or silicon? No, but there's a huge difference between the transistors of today and the biological neuron. The "wet computer" as described is probably not exactly like a biological neuron, but I can assure you it is not like a modern transistor, and that in itself makes this an interesting project.

Prions (1)

Max(10) (1716458) | more than 4 years ago | (#30729498)

From TFA:
"Recent work has shown that when two such lipid layers encounter each other as the cells come into contact, a protein can form a passage between them, allowing chemical signaling molecules to pass. Second, the cells' interiors will play host to what is known as a Belousov-Zhabotinsky or B-Z chemical reaction. Simply put, reactions of this type can be initiated by changing the concentration of the element bromine by a certain threshold amount."

And then some scientists are shocked to find out that prions can evolve.

We'll develop highly advanced "wet computers"... (1)

truparad0x (1589523) | more than 4 years ago | (#30729688)

then we'll: -add AI -make them in the image of ourselves -make them capable of self replication -make a male version and a female version -put them in a perfect, kick@ss new colony...wait a minute...

Is it always wet? (2, Funny)

TooMad (967091) | more than 4 years ago | (#30729690)

Or just when it is turn on?

Re:Is it always wet? (1)

Monkeedude1212 (1560403) | more than 4 years ago | (#30730142)

The power button can be difficult to find. You'll have to search around for a minute or so until the system shows some clear responses. Even then, you'll have to stimulate said power button for a few minutes before it finally boots up, allowing you to do whatever it is you want it to do. Be forewarned though, if you take too long, it will eventually lose power and turn off - however on the other hand if you are too quick it may not want to turn on at all next time. You should exercise extreme caution and timing when using your new wet computer. It may take multiple times to practice the rhythm.

Re:Is it always wet? (1)

ArsonSmith (13997) | more than 4 years ago | (#30730392)

sound's like it'll be bad for surfing porn then, probably be a waist of time and money to even attempt using it.

Re:Is it always wet? (0)

Anonymous Coward | more than 4 years ago | (#30733736)

Perhaps you need to surf porn to get it wet. O_o

Doctor in the House? (2, Funny)

muphin (842524) | more than 4 years ago | (#30729706)

great now when my PC breaks down I have to pay its medical bills too?

Re:Doctor in the House? (1)

ArsonSmith (13997) | more than 4 years ago | (#30730338)

It's OK Obama care will fix it for you.

Easy overclocking (1)

LockeOnLogic (723968) | more than 4 years ago | (#30729972)

Just throw a few ritalin in there, viola! If you want to go for XTREME overclocking you have to use crystal meth, but you risk overheating.

Re:Easy overclocking (0)

Anonymous Coward | more than 4 years ago | (#30730232)

Just throw a few ritalin in there, viola!

Oops, I threw a few Benadryl into my ukulele instead. Now how do I get them to stop rattling around while I'm playing?

Re:Easy overclocking (1)

gstoddart (321705) | more than 4 years ago | (#30730766)

Oops, I threw a few Benadryl into my ukulele instead. Now how do I get them to stop rattling around while I'm playing?

In the case of a ukulele, stop playing. The rattling will stop. ;-)

(I kid, I kid to any of you rabid ukulele fans out there.)

Cheers

Re:Easy overclocking (1)

Hurricane78 (562437) | more than 4 years ago | (#30733066)

Yeah, and as soon as you stop with the Ritalin (e.g. because your wallet is drained empty), your brain will look like this [sandia.gov] , because of what is known as “brain-zaps”.

AI Viruses? (1)

Aksimel (1347591) | more than 4 years ago | (#30729976)

"..it will open up application domains where current IT does not offer any solutions - controlling molecular robots, fine-grained control of chemical assembly, and intelligent drugs that process the chemical signals of the human body and act according to the local biochemical state of the cell." Interesting possibilities abound when you have microscopic computers running around our bodies. Where will we buy the vaccines? Pfizer or Symantec?

Batman reference (1)

Requiem18th (742389) | more than 4 years ago | (#30729980)

I remember watching an episode batman Batman: The Animated Series [wikipedia.org] about an AI developed in wet tissues. I think it was called "Heart of steel" but can't tell...

so-called lipids (1)

vlm (69642) | more than 4 years ago | (#30730174)

"so-called lipids" is a strange phrase. What does that mean? It is a lipid or it is not a lipid.
It's not osama bin laden's secret alias, a value judgment, a marketing campaign, or a trademark.

http://en.wikipedia.org/wiki/Lipid [wikipedia.org]

I realize it was probably written by a journalist, so standards are pretty low, but still...

I will now hit my "so-called submit button" to post this.

Re:so-called lipids (1)

Kozz (7764) | more than 4 years ago | (#30730382)

"so-called lipids" is a strange phrase. What does that mean?

I believe that some journalists, when writing for the proles, will use the phrase "so-called" as a hint to mean, "Look out! Here comes a word or phrase with which you may be unfamiliar!" Stupid usage, I grant you. I don't defend it, but I've seen it before.

Skynet? (1)

phreakincool (975248) | more than 4 years ago | (#30730310)

I guess we're all fucked when it becomes self-aware. Do these scientist even watch the movies?

Simulating Neurons.... (5, Insightful)

jameskojiro (705701) | more than 4 years ago | (#30730480)

Is a lot like building an abacus using 3-D software and then manipulating your 3-D abacus to add 1 plus 3 to get four while chewing away millions of computational cycles...

We need a better way to simulate the effect of a neuron without having to re-create everything down to the last protein and lipid in a nerve cell....

Re:Simulating Neurons.... (1)

dissy (172727) | more than 4 years ago | (#30732898)

Is a lot like building an abacus using 3-D software and then manipulating your 3-D abacus to add 1 plus 3 to get four while chewing away millions of computational cycles...

We need a better way to simulate the effect of a neuron without having to re-create everything down to the last protein and lipid in a nerve cell....

While you are correct, we should probably first focus on a functional prototype, and then we can worry about making it more efficient :}

New meaning for a old expression. (1)

Cr0vv (1223332) | more than 4 years ago | (#30730988)

I guess this gives a new meaning to: "My computer died". blackcrow.

Imagine... (1)

lymond01 (314120) | more than 4 years ago | (#30731522)

...a forest of these interconnected faux neurons.

Then imagine that everyone was blue.

Uh, tech support? My PC isn't workin (1)

mnemotronic (586021) | more than 4 years ago | (#30731542)

I kinda dropped a sponge on it. I squeezed the goop back out, but I think maybe the sponge had some windex and everclear in it ...... yea, it was a pretty good party. The leds are blinking, but they're, like, orange and purple. Is that normal?

Eureka! (1)

cyberworm (710231) | more than 4 years ago | (#30733858)

I'm pretty sure I've seen this on television.
Load More Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Create a Slashdot Account

Loading...