Is the Human Brain the Fastest Cpu

Page 6 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
"The way to make a million dollars is to start a religion."

Hey, Ninja. I think we'd both better be careful the next time we stick our hands in our mailboxes... there might be a Scientologist-placed viper in there... it's been known to happen! 8O
 
Dont forget something, cpu's were made to do aritmetic operations, there are humans that own calculators in aritmetic operations, actually i know this colombian guy Jaime GarcíaSerrano that he bet the calculator, teh calculator took 7 secs to do the operation and the guy have it in 4, he also is trying to get his 6 record guinnes getting 50.000 decimals from the "Pi" all this only with an abacus and his head. here is an site of him its in spanish but u can translate the site. www.jaimegarcia.com
 
If you dabble into the human imagination then anything is possible. :wink: For all we know one of these crazy ideas could possibly be true.

Who says life has to be carbon based.....it could also be silicone based and on other planets you could have life forms based on silicone that use computers with carbon based processors. 8O
 
If you dabble into the human imagination then anything is possible. :wink: For all we know one of these crazy ideas could possibly be true.

Who says life has to be carbon based.....it could also be silicone based and on other planets you could have life forms based on silicone that use computers with carbon based processors. 8O

Well, about crazy ideas possibly being true, from the Wiki again:

The volcanoes that Hubbard specifically mentions in the story (notably Las Palmas and Hawaii) had not formed when Incident II is said to have taken place.

The related Incident I is set four quadrillion years ago, which is nearly 300,000 times the currently accepted age of the Universe of 13.7 billion years.

76 planets of aliens filled with 178 billion aliens per planet totals 13.528 trillion aliens that were packaged and blown up by Xenu. Hubbard did not elaborate on the number of space planes required to transport a population of some 13.5 trillion people. The Douglas DC-8, said to be an exact copy of Xenu's spaceships, seats a maximum of 250 people and has a payload of only around 40–50,000 kg, depending on the specific model. This means that, assuming the Galactic citizens had bodies about the same as humans, and the space planes were the same scale as DC-8s, only about 600 to 700 human-sized frozen bodies could have been transported on each such space plane. To accomplish the deed in a single trip, it would therefore have required around 54.1 billion planes with everyone seated or 19.3 billion planes with frozen bodies packed more efficiently.

Assuming the people were about the same size as humans, 76×178 billion×2 ft³ per alien is 184 cubic miles (766 km³).


As for the silicon beings using carbon computers, if I'm one of them and I'm unknowingly being used by a rockthing to play Crysis, I'm gonna be royally pi$$ed off! :lol:

I dare them to challenge me.

Gabe Cazares Former Clearwater mayor
Jennifer Stewart Victim of rape by a former chief supervisor at the Church of Scientology
Jeremy Perkins Scientologist, schizophrenic, left untreated
Lawrence Wollersheim Former scientologist
Lisa McPherson Died under the care of the Church of Scientology
Maria Pia Gardini Former scientologist, lost millions to Scientology
Mary Florence (Flo) Barnett David Miscavige's mother in law, committed suicide with four bullets from a rifle, three to the chest, one to the head
Quentin Hubbard Son of L. Ron Hubbard, committed suicide
Patrick Vic Committed suicide because he couldn't find enough money for the Purification Rundown
Paulette Cooper Writer, author of Scandal of Scientology
Raul Lopez Mentally impaired Raul Lopez was $1.7 million richer before Scientology
Roxanne Friend Former scientologist, drugged, kidnapped, imprisoned by Scientologists when she attempted to leave
Susan Meister Former scientologist, died from a bullet in her forehead aboard Hubbard's ship

And lots more dead Scientologists.

But yeah, I gotta admit, I'd like nothing better than to face off against Miscavige with one hand tied behind my back and still open up a big can of whoopa$$ with the other hand!!!! :lol:
 
Yes, but who is crazy enough to mess with someone that has close ties to the Haitians in Miami? not even Sci-nuts.

Haitians?

Miami?

8O

Ninja, my buddy, my friend! When is your birthday anyway? Do you like chocolates? iPods? Hot chicks with big booty?

Er... I know which one...

I wonder if Ms. Guerra is busy on your birthday...

😀
 
There is an entire field of computing science effectively stemming from this question and technologies such as Neural Networks that developed from studying the human brain.

The answer that will surprise many of you was that by the time the Von Neumann architecture based processors had hit the power of the 386's it was generally agreed that the (then) modern CPU's had more processing power than the human brain.

However it should be noted at the time it was believed that in effect the brain was a very complex Hoffman network and that has since been disproved in two ways.

1.) Brain cells are not connected to every other brain cell implying fewer connections and thus less processing power.

2.) The temperature & chemistry of the brain itself effects how the cells react to inputs, making nodes in the brain much more complex than those in a Hoffman style network.

The thing is the human brain is engineered in a totally different way to a CPU, but effectively each cell in the human nervous system has multiple analogue inputs & outputs connecting to other cells in the system.

When the combined analogue inputs exceed a certain threshold the cell is activated and fires an output. The thresholds and outputs are the really interesting bits and depending on what model(s) you agree with can radically effect calculations on just how powerful the human nervous system is.

But the problem comparing the two is much more fundamental, we almost understand how the brain is engineered (in terms of nodes) and almost understand how to train a node or even groups of nodes. Computers have even become powerful enough that they can model networks as complex as the human brain in real-time on a very different architecture.

The problem is we have no idea what the thresholds & outputs of the human brain are, the whole secret to the human brain is not the hardware (the nervous system) that's actually not that powerful and relatively simple. The problem is the software it runs (the thresholds & outputs) evolving a neural network that complex involves immense amounts of CPU power and worse still we still have no idea exactly how the software of the human brain evolved and thus how to write something similar (remember building the hardware is the easy bit we know how the hardware evolved).

Neural networks are really, really powerful at doing some things despite having a low speed try solving a traveling salesman problem on traditional software / hardware. Give it a go the problem just keeps growing exponentially as you add locations but throw a neural network at it and even though its much simpler it will deal (rather than solve) with the problem much more effectively.

P.S. for more on neural networks start here http://en.wikipedia.org/wiki/Neural_network#Types_of_models
 
The Wiki - Read up on the Singularity

There are some very impressive theories about what will happen as computer intelligence approaches human intelligence. Mix in the possibility that we may be able to use nanotech to augment our own intelligence and you have some interesting times ahead.

Some food for thought:

If we develop nanotech (or any technology) that can speed the data rate within our brain by a factor of 2, so now the brain is capable of performing twice the functions in the same period, what effect does that have on our reality?

Does the rest of the world seem slower to you since you are processing the same amount of external input at twice the speed? Do you just feel twice as smart?

If your brain were "mapped" and uploaded to silicone in a perfect replica body would you even know it?
 
heh. For anyone that may have noticed, my avatar is the most basic weighted algorithm used in neural networks.

It's actually pretty simple: each neuron represents a single unit of statistical analysis; a single unique mathematical algorithm, if you will. Each neuron has multiple inputs and a single ouput. Like A+B+C+D = Y. That output is usually wired to thousands of other neurons.

A neuron could be compared to an ALU. The difference is that neurons are generally function-specific, whereas an ALU is general-purpose. This means that the ALU will be used for many different types of functions, where the neruon may only react to, say, motor functions.

You can compare the brain to DX-9, and the CPU is DX-10. DX-9 cards have pixel shaders and vertex-shaders; they only operate when given specific tasks. DX-10 has unified-shaders, meaning they operate whenever there's a generic workload.

For anyone that's done any serious work in AI, you'd know that there's actually a good reason for this. When a single neuron fires, it may trigger other neurons to fire. The other neurons will fire based on (1) their input and (2) the amount of oxygen they have available. Neurons won't fire if they don't have enough oxygen, provided by blood, hence the function of MRI's. This allows a human to 'maintain a train of thought.'

Well, I could go on all day, but really you're comparing apples to oranges. The problem is that in a fraction of a second, you may fire 3,000 neurons simultaneously, representing an incalculable number of statistical computations, where a CPU is only capable of processing a finite number of calculations per second.

In short, a brain is 'slow' by design, so the answer is "No." However I'm sure there are times when the number of raw calculations that occur in a split second in the human brain would cripple even the beefiest of server farms.
 
Well some of the software is in our DNA that is passed along from our parents. That software will determine a lot of things including the make up of our brain and its intellectual capacity.

So the brain comes with some basic software but is fairly open to any kind of input. Then when we are born we are basically programed by our parents, environments so the software is actually inputed via our environment.

So if you were raised by chimps you would know nothing more than what a chimp knows. The brain is not that impressive unless it is programed correctly and some brains are easier to program than others. :wink:

Scott Adams who wrote the Dilbert Principle baiscally said that people really are not that smart and that the world we live in now was basically created by a small number of geniuses. Think about it how do you know what you know? Did you invent it or create it? Could you have created the printing press? The light bulb? The computer? Think about it there is a lot of truth to it. Things we use every day do we really know how they work?
 
Here is something else we can throw into our equation/comparison

Quantum computer to debut next week

http://www.techworld.com/opsys/news/index.cfm?newsID=7972&pagtype=all


I wonder if this will be the ENIAC of quantum computers or another cold fusion fiasco...

Yeah, like I don't have enough problems with computers, now we have one that will stop working IF YOU LOOK AT IT!

Doesn't it strike anyone else as absurdly funny how badly physics misunderstands quantum mechanics?
 
Can't resist a comment on how much memory the brain has, there are those two would argue (including myself) that the total memory capacity of the brain is, wait for it 0.

The brain / nural networks just don't work like that. Memory / storage is a more Von Neumann architecture pinciple where you are taking and returning instructions.

In the nural network there is just the current liquid state and thresholds, now you could argue that this is a form of memory and given that you have analogue inputs, with analogue timings (debatable) and analogue distance the memory capacity of a single input on a nueron could be considered astronomical.

Psychological testing has suggested that actually we don't remember as much as we think we do, memories in the brain are forms of linkages where we relate ideas (hence why we get confused / cross ideas) in reality its not single ideas but inter-relationships that we remember (which given the architecture of the brain makes allot of sense).

Look at your desktop background, something you see and remember probably every day, go next door and draw it. Even if you were are a great artist and could exactly re-produce what you see in your mind the picture you drew would not match your background, things would have subtally moved, colours etc changed and if you could have somehow indexed everything you had seen in you life you would find aspects of other things you had seen. By the way some people initially see this as a flaw in design but this is actually the greatest strength of the brain.

Sure we remember things, lots of things, telephone numbers, how to get to places, how to do a million things but in many cases it is processes that we have combined & adapt to the task that are built in rather than exact memory and often even the things we remember will change every time we do them.

I'm not explaining this very well but in effect because the brain is in a constant dyanmic state of flux it is always processing but never storing so the images, facts etc that you are remembering you have been remembering / processing constantly since you learnt / experienced / adapted them.

Being honest I'm not very good at explaining this I did about 2 years of AI & Nural and adaptaive at uni and lived with a bunch of psychologists and would talk for endless hours about the nature of the brain and our understanding of it and the only thing we ever really managed to reach agreement on was that we didn't understand it or its nature.

P.S. warning to anyone upfront, nural and adaptive technologies are tough, invole allot of mathematics (neither of which bothered me) but the very nature of the science causes you to question things which you shouldn't and will really, really, really screw with your head.
 
:lol: :lol: :lol:

BTW, if I understand correctly, you are a long time forum member that has been away for awhile, I just wanted to say welcome back !

Sometimes the forums get a little dry and flamed into oblivion, a little humour goes a long way. :)
 
:lol: :lol: :lol:

BTW, if I understand correctly, you are a long time forum member that has been away for awhile, I just wanted to say welcome back !

Sometimes the forums get a little dry and flamed into oblivion, a little humour goes a long way. :)

When I was first hanging around here, we were all waiting with bated breath and arguing over the possible stratospheric benchmarks of the upcoming, exciting, zoomy 80286! :lol:

Thanks for the props! I 'preciate it!
 
I dont think computer will match any animal on earth. Movement, sound, taste, feel, Feed it self. feel danger, see danger. Or adapt in differnt climents. We dont see laptops moving around the house to get shade to cool off or protect the lsd Monitor from sunlight. All there is a fan. We dont see them cleaning them self of dust, Basicly taking a crap.

Robots can move, microphones hear in almost exactly the same way we do, chemecal detectors are are present in EVERY home (smoke and carbon monoxide detectors), if machines can't "feel" then how do you think you managed to type your response? Your keyboard "felt" you depressing the keys in a very similar way that we sense pressure. SpeedFan can sense warning conditions and react to them, Samsung has made working autonomous machine gun turrets that track and fire on human targets. Laptops do respond to heat and light. Fans alter their speed based on heat and there are monitors that change brightness based on light sensors. There are also self-cleaning air filters.

You're going to have to try harder than that. There is no mundane task that hasn't already been accomplished with technology. Some of them haven't been accomplished very well, but it's all been done. If you think something as mundane as your tongue is what makes you a special and valuable fixture on this planet you might be one of the people we'll replace with robots first 😛

There are strong "humans rule" arguments to be made concerning creative and inventive abilities and you could even argue that software cannot have "emotions" (but that's a little difficult to quantify and not really something that would make humans better anyway no matter how many cheesy SciFi movies you've watched) no need to bring up horribly mundane things like taste and extretory abilities. You could argue that the average human being is a lot better at them compared to any system currently available, but saying that technology can't do them at all is just silly.
 
i think this would be the best place to discuss this

in another thread the point was raised that the fastest CPU is the Human Brain not wanting to hijack a thread i thought i would post a new one

The Human Brain is the Fastest Cpu

now thats open to debate. seeing as a brain acts as both a hard drive and CPU also.

also what is 928346*98236?
im sure the computer can do that faster than the human brain hell try doing what super pi does in 30s in your brain.

i would suggest that a human brain is more like a 100 core processor with a lot of them dedicated to individual functions that wouldnt be much good at doing other things (cpu cant act as a Gpu)

what makes something the fastest CPU?

Think of it this way, Core 2 Duo 6400 is a dual math processor at 2.13 Ghz that can multi-task has an x86 based instruction set. Your brain is a multi-core word processor, math processor, GPU and more that probably runs at 2 to 3 hz, does not multi task, has a has a more sophisticated instruction set. In other words the CPU is faster that the Human Brain but the Human Brain has more capabilities. The Human Brain is designed to do something completley different than the CPU. It is like comparing the CPU to GPU.
 
i think this would be the best place to discuss this

in another thread the point was raised that the fastest CPU is the Human Brain not wanting to hijack a thread i thought i would post a new one

The Human Brain is the Fastest Cpu

now thats open to debate. seeing as a brain acts as both a hard drive and CPU also.

also what is 928346*98236?
im sure the computer can do that faster than the human brain hell try doing what super pi does in 30s in your brain.

i would suggest that a human brain is more like a 100 core processor with a lot of them dedicated to individual functions that wouldnt be much good at doing other things (cpu cant act as a Gpu)

what makes something the fastest CPU?

Think of it this way, Core 2 Duo 6400 is a dual math processor at 2.13 Ghz that can multi-task has an x86 based instruction set. Your brain is a multi-core word processor, math processor, GPU and more that probably runs at .01 to .03 hz with its own throttling according to tiredness, does not multi task efficently and has a more sophisticated instruction set. In other words the CPU is faster that the Human Brain but the Human Brain has more capabilities. The Human Brain is designed to do something completley different than the CPU. It is like comparing the CPU to GPU it is a stupid idea.
 
does not multi task

I'd like to disagree with this statement. While on the surface we might not be able to multitask efficiently, our brains seem to be optimized to allowed a small percentage to continue working on problems while one primary process takes up most of the cognitive abilities.

Then there are all those backround processes running, which take up something like 80% of things. If half of that were switched over to cognition with an external processor taking over those things like breathing, digestion, heart rate, blood production, etc., the anyone would instantly become a genius. Well, likely not, since each part of the brain is probably devoted to a specific duty.
 
Status
Not open for further replies.