Geeks Weigh In: Does a Human Think Faster Than a Computer?

supercomputer   Geeks Weigh In: Does a Human Think Faster Than a Computer?While many people stereotype geeks as only being interested in using the computer all day, the truth is that a geek is actually a person who often contemplates many of the deeper questions of the universe while busy installing the coolest new add-ons to Firefox or tweaking their mobile phone so that they can control it from their desktop. One of the universal debates many geeks have centers around an important question that involves neurobiology and the science of artificial intelligence, and that question is – Does a human think faster than a computer?

What a question. Just think of the necessary evidence that one would need to produce in order to prove, or disprove, that statement. In fact, what is the question about really? Is it whether a human brain or a computer is faster, or is it which form of information processing is better? Is it even a fair comparison? Today, I’d like to engage MakeUseOf readers into a debate on this subject by first providing my own take – and then asking for yours.

The Question: Does a Human Think Faster Than a Computer?

The question itself represents the fallacy of how people think about computers. When a person uses a computer, if it’s slow then it’s junk. But there are certainly other factors to consider when examining intelligence – what about image recognition, language recognition, multi-tasking capabilities or self-learning and self-healing features?

neuralnetwork   Geeks Weigh In: Does a Human Think Faster Than a Computer?

First, to partially answer the “speed” question we need to examine data transmission. In the Hartford Examiner, writer Joy Casad answers the question, “How fast is a thought” by describing the chemical/biological propagation of “thinking” neurons before getting to the point in the final paragraph – these neurons transmit signals at 0.5 milliseconds. That’s pretty fast!

In 2006, the fastest reported fiber optic transmission rate was 2.56 terabits a second. Okay, but a bit is nothing more than a zero and a one. Well – the current state of the art is the cutting edge subatomic technology created by Stanford researchers representing one bit with 35 electrons, or 35,000,000,000 electrons a millisecond. Due to the fact that axon/neuron electrical transmission depends on the chemical and biological environment it is in, data transmission of one neuron is actually millions of times slower than the fastest electrical transmission rates over copper electrical wire, and even slower compared to fiber optics. Score one for computers.

What About Processing Power?

The question of processing is a tricky one. According to the Top500 list of super computers, the fastest one as of 2009 is the RoadRunner BladeCenter at 12.8 GFlops (floating point operations per second).

bladecenter   Geeks Weigh In: Does a Human Think Faster Than a Computer?

A GFlop represents a billion operations per second. Now, you’re thinking of that Monday morning in class when your professor asked you to perform a simple calculation and your mind went blank. You’re ready to chalk up another point to computers, right? Wrong.

While the transmission of electrical impulses may be slower in the brain than over wire, the processing power of the brain is represented by not one, but thousands of processors backed into one major super computer. One example is the retina, which is sort of like your computer web cam, in that it transmits light (images) to the brain for processing. Except the retina itself has its own processing power, sort of like a subprocessor – 100 million neurons packed into a one centimeter by one millimeter space.

humaneye   Geeks Weigh In: Does a Human Think Faster Than a Computer?

This stunning little processor is capable of processing ten images, each of about a million light points, every single second. Not only that, the data isn’t transmitted over a single fiber of nerve cells, but over a cable to the brain made up of a million of these fibers, all transmitting bits of data at the same time in parallel. If you multiply the processing power of this volume of neurons by the overall size of the average 1,500 cubic cm human brain, the overall processing power of the brain is about 100 million, million operations per second. For those of you who are trying to do the math with your super computer brain – that’s over 100,000 times more processing power than today’s cutting-edge super computer.

Image and Language Recognition, Learning and Common Sense

If our brains are such super computers, then why do we feel so dense and so slow sometimes? I don’t know about you, but I’m horrible at doing calculations in my head. The problem is that people think of computers only in terms of how many calculations it can do per second. The truth is, when it comes to intelligence there’s so much more to process than calculations alone. How do you calculate what the tone of someone’s voice implies they are really saying?  How do you calculate the irony of a joke that, when taken literally, makes no sense at all? This is where the true power of the human brain makes itself known.

jokemilk   Geeks Weigh In: Does a Human Think Faster Than a Computer?Have you ever had a friend who was such a genius that they could perform the most astounding calculations in their head, or they could fathom the most complex equations or problems imaginable – yet when faced with the simplest common-sense joke, they just didn’t get it? This is the major difference between a human brain and a computer.

Author Gary Marcus writes, in his book on the human mind that, “The fundamental difference between computers and the human mind is in the basic organization of memory.”

What he means is that a computer organizes information in a logical way. To retrieve data, the computer uses logical storage locations. A human brain, on the other hand, remembers where information is stored based on cues. Those cues are other pieces of information or memories connected to the information you need to retrieve. This means that the human mind can connect an almost unlimited number of concepts in a variety of ways, and then sometimes disconnect or recreate connections based on new information. This allows the human to step outside the boundaries of what has already been learned – leading to new art and new inventions that are the trademark of the human race.

There are a lot of other ways the human mind blows computers away – it can self repair itself, it can produce chemical reactions within its host body to induce instinctive reactions and protect itself from danger, it can handle every last function required to operate the machine of the human body while simultaneously processing information from outside that body, and most importantly it can continue learning and building new connections within that contextual storage array in ways that seem infinite.

In short, the answer to the question “Does a human think faster than a computer?” is yes. And it can also do a whole lot more than that.

Geeks out there – weigh in with your opinion in the comments section below!

Image Credits: cbowns

The comments were closed because the article is more than 180 days old.

If you have any questions related to what's mentioned in the article or need help with any computer issue, ask it on MakeUseOf Answers—We and our community will be more than happy to help.

34 Comments -

0 votes

roger

this afternoon i used my computer to download 180Mb of music. i think the time to make comparisons is when my computer can not only do the downloading but can make the kind of value judgements i make regarding which material to download.

0 votes

Chuck thomas

Wow, no way dude that is crazy!

RT

0 votes

Nick

Great Article. I really does make you think! Do you think the processing power from a true genius would be different than a average person?

They say someday computers will be smarter than us, but I just hope not…cause that would be bad IMO.

0 votes

J-Bo

Something to think about in that regard is that the brain’s processing power isn’t bent towards intelligence alone. For example, think about NFL running backs or other professional athletes. It takes a lot of brain power to scan the field, make a decision, make continuous predictions, and actually move your body that quickly in accordance with your integrated observations. Anyway, yeah, I’d like to add athletes into that comparison as well.

0 votes

Nick

That is very true indeed. The brain is just so amazing and who knows what else it is capable of doing.

0 votes

Lewis

The IBM Roadrunner runs at MUCH more than 12.8GFlops. Each of its CPUs runs at 12.8GFlops and when you add them all up you get almost 1.5petaflops….thats 1.5 thousand million million, 15x the number in the article for humans.

0 votes

Lewis

Almost forgot – the Stanford researchers’ findings show 35 bits per electron, not 35 electrons per bit. Off by a factor of 1225.

0 votes

Jay

Lewis, being precise and being a dork are not mutually exclusive

0 votes

Ryan Dube

Lewis – thanks for the correction, although on the scale that we’re talking about here, I’m not sure that a “factor of 1225″ makes a heck of a lot of difference…lol. But, thank you for keeping the numbers perfect. :)

0 votes

David

Lewis, that sounds much better! lol. I saw that number and I was thinking… I think the ps3 can do better than 12.8 Gflops. It has 7 processors or so and each go at 3 ghertz.

0 votes

billy

Lewis & David, have you never worked with computers before? You don’t add each processor together. It’s 12.8 GFflops no matter how many processors there are. The 12.8 Gflops is shared between all the processors. NOT added together.

0 votes

curtis

i lol’d at them.

0 votes

Ryan Dube

This is true – although the concept of parallel processing (in the way the brain does it) is an interesting concept isn’t it? The brain can run the many control systems within the human body that we don’t consciously even know we’re “thinking” about – at the same time it’s taking input from every sense – touch, sight, smell, taste – and at the same time additionally contemplating the ultimate meaning of life and the universe. :)

0 votes

Lewis

Yes, I’ve worked with computers before. My aging 9800GT GPU alone will do much more than 12.8 GFlops – I certainly hope the most powerful supercomputer in the world can beat a $75 video card.

http://en.wikipedia.org/wiki/IBM_Roadrunner

“In November 2008, it reached a top performance of 1.456 petaflops, retaining its top spot in the TOP500 list.”

Also, @J-bo; remember TP from G4ums? *waves*

0 votes

Fisher

One thing you could have brought up would be data storage. It seems amazing to me that i can remember exactly what I did in a water polo game four years ago, but likely wouldn’t be able to remember what I had for dinner just two days ago for my life. On the other hand my computer remembers any and everything it wants to until it suddenly decides that it no longer needs that data and deletes it forever.

0 votes

JoeT

Didnt know Roadrunner only did 12GFlops.
Though it was alot closer to 1.5 petaflops…

0 votes

Ryan Dube

Thanks for the correction guys – I noted the processor not the combined Pflops…but even with the extraordinary ability of computers to execute umpteen thousands of math computations per second – the real question to discuss here is whether that constitutes greater “intelligence” than the human brain. So – does it?

0 votes

Will L

I don’t care what ‘geeks’ say, the computer has been being developed by man over the past how many decades… the mammalian homo sapien brain represents millions of years in neurological evolution, and the biological technology that correlates with our brain and makes up even our most basic fundamental of life as DNA is billions of years in the making. Even with our most powerful tech we still barely understand many aspects of the brain.

Chock that up to the brain.

0 votes

Jeroonk

The Roadrunner has a parallel capability of hitting 1.5PFlops. Which is about a factor 117,000 over 12.8 GFlops.
Yes, you can’t just add parallel capability together, but same goes for the brain!

You can’t just add up the speed of parallel neuron processing.

0 votes

Ryan Dube

Guys…thanks for the correction on the Roadrunner, I appreciate you guys setting that straight – but what I’d *really* love to hear is your opinion on the brain vs. computer debate? Do you have one?

0 votes

Mistakes Girls Make

Computers are faster at processing information but not retrieving it.

0 votes

seif sallam

so the conclusion from that is:
our unconscious mind is faster than the fastest super-computer, but our conscious mind is very slow compared to a super-computer

0 votes

BobT

The real issue is how does self awareness and ability to perceive meaning arise? These are all based on subconscious processes that when operating together at a certain threshold produce concsciousness. No one remembers being conscious a young baby in their mothers womb or for the first few years after birth and yet you were alive learning things.

Computers will eventually do the things we do and we will teach them to understand the things we do, but whether or not htey will be aware of their existence as we are I doubt until we unlock the mystery of how the brain produces consciousness, I imagine a kind of hybrid will emerge for true AI, there are probably trade offs between different kinds of computation, i.e. biological is better then traditional at certain classes of tasks, while traditional computers are better then people at some tasks.

The difference is that the computer is not aware that it even exists or is doing anything. But I also imagine that many animals and insects are not self aware as we are either. There is a threshold where unconscious biological systems give rise to true self awareness. What it’s mechanisms are are not yet known and will require years of work.

0 votes

Ryan Dube

What a great & thorough response BobT – I enjoyed reading this.

0 votes

sam

I’m laughing my ass off that you actually thought the most powerful computer in the world was 12.8 GFlops…the new Radeon 5870 does over 2.5 TFlops.

0 votes

Ryan Dube

Thanks – the processor is actually 12.8 GFlops as noted on any of the specs you read on it – so while it’s cute that (in light of all of the reader comments above laying out the true overall processing power of the Roadrunner) you can laugh your ass off – why don’t you offer a thoughtful opinion on the larger issue while you’re at it?

If the Roadrunner does have a greater ability to process more operations/computations per second than the human brain, does that make it more “intelligent”?

0 votes

Marcelo

Any geek who says computers are faster than a human in terms of processing power, seriously needs a biology lesson. Any other answer is based on ignorance that their users have when calling them for IT work.. The human body is just TOO complex, just imagine all the thins your body just when you are in “Sleep mode”. Nuff said.. When computers or the like do reach that state, which is completely possible, then we are deep trouble..

0 votes

RegBes

Processing Power != Intelligence.

You need to define intelligence clearly for a definitive answer. In basic intelligence tests like games of chess and go computers have been gaining and surpassed all but the very best human players of these games.

A better test of intelligence is the Turning Test http://en.wikipedia.org/wiki/Turning_Test

The programming languages and the programs that have been written to make computers (appear?) intelligent, is where the rubber meets the road and where this question will be answered not in processing power.

0 votes

drawforjoy

Very interesting article. Anyway, our mind is better and funnier.

0 votes

Gir

Disclaimer: I am not, by any means an expert of any sort, but just a student that does a bit of reading from time to time, and it has been a year or two since I read these types of articles.

It is my understanding that our brain cells do not only process one bit, but that they process one or more q-bit/s(quantum bit/s). I believe(from my understanding) that comparing bits to q-bits or Bytes to q-Bytes, etc, is again like comparing apples to oranges.

For a little more research, you can go to:
quantumconsciousness.org
or look into: quantum consciousness, microtubules, look into some quantum theory, or possibly read about quantum computers

I really enjoyed reading this article, but mostly I just wanted to stir the pot a bit. I don’t care if any or all information I have presented is disbelieved or thrown out, I just thought it deserved stating.

0 votes

Soph

You’re comparing a flop to a neuron firing as your benchmark “unit” and then making comparisons in terms of these. It’s too apples-to-oranges a comparison to give any kind of meaningful answer.
Is a neuron like a processor? It evaluates input and computes an output, so it is somewhat processor-like, but it’s output is just a 1-bit, yes/no, fire/no-fire, answer. Even the most active neurons in the brain fire around 10-100 times per second (perhaps as much as 300-500). 1k per second or so is probably the maximum firing rate, and if you down some ecstasy with a can of redbull, snort some coke, smoke some meth, strap on your nitrous mask and have sex with a supermodel while riding a rollercoaster through a fireworks display at a rock concert, you might get some of your neurons to fire that fast. You’ll also cause quite a bit of permanent brain damage.
A neuron then is somewhat like a slow, 1-bit processor. It would appear rather pathetic compared to even the rather run of the mill processors that control things like, say, an elevator, or the electrical timing of a modern car’s ignition system. We now have 64, 128 and 256-bit procossesors, operating many millions and billions times faster. By these standards a neuron sucks and the brain is pretty lame. But flops is a lousy proxy for intelligence. Ramp your brain up to a million times its normal speed and you’d think a lot faster, but you wouldn’t be any smarter. If you didn’t understand something with a 1khtz brain, you still wouldn’t understand it with a 1petahtz brain. On the plus side it would only take one trillionth as long to realize that you didn’t understand it.
There are around 10 billion neurons in an adult human brain (Specifically there are around 10 billion cortical pyramidal cells, which are believed to be the main cells responsible for cognition, and billions more glial and other support cells). The better analogy here might be to the transistor. Transistors can be used as switches, like the on/off firing of a neuron, and they can be used as amplifiers, a slightly more complicated process (though not nearly as complicated as the process by which a neuron fires). The quad-core Itanium processor under development by Intel will have around 2 billion transistors, or about one fifth the number of neurons in a human brain. By the imperfect comparisons used thus far one could say that it is one fifth as “powerful” as a human brain, though many million times faster. Then again, as a quad-core processor, it might be more akin to four rat brains wired together and floating in a bucket of liquid amphetamines.
I would be very surprised if there weren’t super computers out there that didn’t already have more than 10 billion transistors, and operate billions of times faster than the brain. But, where the brain really excels is in networking. Typical transistors have just two inputs and one output, a neuron has thousands of inputs and thousands of outputs. Synapse–the points of connection between neurons–number around 100 trillion, or 100,000,000,000,000. The mylenated axons, or “white matter”, that branches out from the main body of each neuron, aka “gray matter”, is so prevalent in the brain that if every axon of a 20 year old adult male were laid end to end it would stretch more than 120,000 miles, enough to circle the Earth one and a half times. That’s a lot of ethernet cable packed inside the average noggin’.
All these comparisons are silly. Brains and computers operate on fundamentally different architectures. Computers operate in serial, a sequence of steps operating on an external set of data. Brains operate in parallel, and the data storage and processing take place in the same place. If it can be broken down into a series of discrete steps that need to be applied over and over, or if it involves large data sets, use a computer. If it involves pattern matching, a brain works better.

0 votes

Ryan Dube

Soph – you’ve just pulled off an excellent comparison, and a great article in itself. Thanks for writing it and offering your opinion, I enjoyed the read!

0 votes

Doctor Fonz

Your the topic on processing power is *very* incorrect.

You mention that the 2009 Supercomputer ‘Roadrunner’ is ticking over at 12 GFlops – heck, my graphics card is running at over 1000 GFlops, aka 1 TFlop.

I think you’ll find that Roadrunner actually has the processing power of 1.1 MILLION GFlops, also known as PFlops or Petaflops.

This flaws this plus point to humans I’m afraid :-( …and thats not taking into consideration distributed supercomputers made up of millions of smaller nodes. Folding @ Home peaks at over 7 Petaflops, making it six times faster than Roadrunner!

Evidence of Roadrunner’s speed at the following URL :-
http://www.top500.org/system/performance/9707

0 votes

Call Center Guy

Ok now I can’t look at anything without thinking about how fast my brain is processing and calculating these images.

As for the debate, until computers can smell and feel, while doing other things such as sitting up straight, feeling hunger, and craving for a smoke while sitting in the office reading blogs rather than doing work, human brain > fastest computer.