A new kind of thinking needed?

[GreatBigBore]

Up front: I'm a total layperson with no formal scientific training. I'm also a rational human being who doesn't believe in mumbo-jumbo but can allow for there to be plenty of counterintuitive reality that we haven't yet discovered. No Bigfoot or Nessie talk, but, you know, open-minded.

So I recently viewed a moving speech by Carolyn Porco (on youtube at watch?v=qGSv-uZCOyY). She is very fond of Galileo, and credits him, more or less, with the invention of the scientific method. I'm not sure I can believe this to be true, as it seems obvious to me that the scientific method (at least the basic idea of testing one's opinions against reality) is hard-wired into our behavior (just watch your kids), and it seems staggering to me that it took us so long to formally articulate the concept. But ignoring that point for the moment, it occurred to me that until Galileo, human reason had really gone as far as it could go, and a new way of thinking was required.

It has always bothered me that the extremes of our knowledge (e.g., quantum and relativistic phenomena) are so out of whack with our intuitions. It's a good point that we evolved in this middle order of magnitude so we lack good intuitions at the extremes. But science has been challenging our intuitions for centuries--quantum mechanics and relativity weren't the first surprises, after all. Yet every time our intuitions have been challenged, we've quite readily assimilated our new discoveries into our thinking, and without even realizing it we absorbed the surprises into our intuitions. But extreme phenomena still challenge our intuitions--we can't seem to assimilate them. Hence the craziness of string theory, for example.

It seems to me that perhaps we need a new cognitive tool in order to take the next big step in human understanding. The scientific method is an indispensable tool--we'll always need it--but we need a new, additional tool. A new way of thinking, some mode of thought that we can train ourselves to use that will allow us to have a better, and intuitive understanding of the extremes.

Having such thoughts rolling around in my mind, I discovered Daniel Tammet, who has a rare and extreme form of high-functioning autistic savant syndrome. He can do amazing math and language stuff, without trickery. It's just that his mind works differently. And I guess he has written a book suggesting that we can train ourselves to think like he does.

So the big question behind this long speech is this: has anyone thought to discuss the extremes of our physics with any of the Daniel Tammets of the world, to see if they have anything interesting to tell us? Perhaps some patterns that the rest of us aren't seeing?

 

[Chronos]

Intuitive constructs are insufficient to supplant mathematical formalisms. In other words, you belong in the IR submission forum. You might be right but you are appealing to the wrong audience.

 

[Chalnoth]

It seems to me that perhaps we need a new cognitive tool in order to take the next big step in human understanding. The scientific method is an indispensable tool--we'll always need it--but we need a new, additional tool. A new way of thinking, some mode of thought that we can train ourselves to use that will allow us to have a better, and intuitive understanding of the extremes.

Except an intuitive understanding of the world around us really isn't necessary to science. It can be kind of appealing, I suppose, but it isn't necessary. And our brains are probably physically incapable of it, just as we are physically incapable of visualizing four dimensions.

 

[GreatBigBore]

Except an intuitive understanding of the world around us really isn't necessary to science. It can be kind of appealing, I suppose, but it isn't necessary. And our brains are probably physically incapable of it, just as we are physically incapable of visualizing four dimensions.

Intuitive constructs are insufficient to supplant mathematical formalisms.

I'm not saying that an intuitive understanding is necessary, and I am in no way suggesting that we do away with the tools that have brought us this far, such as mathematics. I'm using our lack of an intuitive model (your example is perfect: our inability to intuitively grasp a fourth dimension) as a suggestion that we're coming close to the limits of our existing mental tools. I'm saying that perhaps in the same way that Galileo's philosophical approach (test your opinions) was an extremely useful (and at the time novel) tool in advancing our understanding, we might be at a point where we need some new models, some new ways of thinking about things, some new math.

You might be right but you are appealing to the wrong audience.

I'm not appealing to an audience. I'm not promoting an agenda. I am simply asking, after having given some background on where the question is coming from, whether anyone on this forum knows whether any physicists have talked with any of the savants of the world to see whether their unique ways of visualizing mathematics and other symbolic systems could be helpful in advancing our understanding.

 

[twofish-quant]

She is very fond of Galileo, and credits him, more or less, with the invention of the scientific method.

I've read people that are quite critical of Galileo, and suffice to say that things are a lot more complex than they first seem, and as with most things historical, it's better if you read some competing narratives to see what really happened.

I'm not sure I can believe this to be true, as it seems obvious to me that the scientific method (at least the basic idea of testing one's opinions against reality) is hard-wired into our behavior (just watch your kids), and it seems staggering to me that it took us so long to formally articulate the concept.

It's pretty obvious to me that the scientific method is *not* hard-wired into our behavior, since most people get it wrong. It's also worth noting that most of our current ideas on what science *is* date from the 1950's. Thomas Kuhn wrote his book in the 1950's, and Popper's ideas came from the 1920's. If you asked a scientist in the mid-19th century and one today on what they are doing, you probably will end up with very different answers.

But extreme phenomena still challenge our intuitions--we can't seem to assimilate them. Hence the craziness of string theory, for example.

1) String theory might be wrong. From a physics point of view, I'd consider it a massive colossal wild goose chase.

2) There are people that *do* have deep intuition about string theory. It's all a matter of staring at something every day for several years, and after a while you develop a "gut feeling" for how things work or don't work.

It seems to me that perhaps we need a new cognitive tool in order to take the next big step in human understanding.

Science usually doesn't work with big steps. It works with lots of little steps. The problem is that the reality is messy, and if you take huge big leaps without data, you are likely to end up not only being wrong, but unable to figure out that you are wrong.

Personally, I've found that things like google, wiki and mathematica are extremely research tools because they make it easier to teach mathematical concepts. A lot of the problem with physics is that you don't have the ability to "play" with objects, and if you have a computer program that let's you play with 4 dimensional hyperbolic spaces the same way that 3-year olds play with blocks, it's much easier to develop an intuition about how things work.

I discovered Daniel Tammet, who has a rare and extreme form of high-functioning autistic savant syndrome. He can do amazing math and language stuff, without trickery. It's just that his mind works differently. And I guess he has written a book suggesting that we can train ourselves to think like he does.

One curious thing though is that some impressive mental feats turn out to be pretty useless in physics and mathematics. The ability to memorize Pi to 22514 digits turns out to be an interesting trick, but it's something that is completely useless in the sciences, because you can log into google and end up with pi calculated to 100000 digits.

So the big question behind this long speech is this: has anyone thought to discuss the extremes of our physics with any of the Daniel Tammets of the world, to see if they have anything interesting to tell us? Perhaps some patterns that the rest of us aren't seeing?

Already done. If you talk to a mathematician or a mathematical physicist, then you'll find that there is no huge shortage of people that have very deep intuition about mathematical structures. One problem is that to even *communicate* what is going on requires about a decade of effort. A lot of higher mathematics involves getting to the point where you can write down what you are seeing in your mind.

The other thing is that the skills that creates a great mathematician often makes for a lousy physicist and vice versa.

 

[twofish-quant]

our inability to intuitively grasp a fourth dimension) as a suggestion that we're coming close to the limits of our existing mental tools.

Who is "we"? Most physicists don't have any problem intuitively grasping fourth dimensional spaces. In order to do any special relativity, you have to have intuitive understanding of Minkowski spaces, and it's not hard to develop this sort of intuition when dealing with flat Minkowski spaces.

Giving people an intuition about Minkowski and infinite dimensional Hilbert spaces is something that's pretty essential for any undergraduate physics curriculum, and people have worked out the techniques for doing that. Dealing with that is kindergarten stuff.

we might be at a point where we need some new models, some new ways of thinking about things, some new math.

Do you have something specific in mind? One problem with talking about "new stuff" is that you need to be pretty familiar with the "old stuff" to know what's missing.

I am simply asking, after having given some background on where the question is coming from, whether anyone on this forum knows whether any physicists have talked with any of the savants of the world to see whether their unique ways of visualizing mathematics and other symbolic systems could be helpful in advancing our understanding.

I think that a pretty large fraction of theoretical physicists and mathematicians *are* high functioning savants with pretty unique ways of visualizing mathematics. One thing about a lot of mental abilities is that they fall on a spectrum with people at one end able to "fake normal" and people at the other end completely unable to function socially.

Personally, I think that there are some interesting cognitive science things to be explored, but it's really tough to do this.

 

[GreatBigBore]

Who is "we"? Most physicists don't have any problem intuitively grasping fourth dimensional spaces...

Your response is making me see my own subtext. What I really want is to be involved, at least as an informed spectator, in the scientific conversation, but the conversation is getting away from me as a layperson. I jumped at Tammet's mesmerizing description of the way he traverses symbolic landscapes as a possible way to make the conversation more accessible to me and my ilk.

I think that a pretty large fraction of theoretical physicists and mathematicians *are* high functioning savants...

Hilariously true; also applies to us engineers, although I'm not sure how many of us are high-functioning.

 

[Chalnoth]

What I really want is to be involved, at least as an informed spectator, in the scientific conversation, but the conversation is getting away from me as a layperson.

Well, unfortunately, that really isn't easy. In much of modern science, we have now moved so far away from the realm of the ordinary, away from the world which our brain evolved to understand, that we just aren't equipped to understand these new facts.

This means that we have to supplant our ability to intuitively understand with mathematics, and that, sadly, requires a lot of study to get right.

Now, granted, once scientists have discovered some interesting new fact, it is a very good idea to try to explain that fact in plain language that everybody can at least begin to grasp, instead of just leaving it in mathematics. But the plain language translation of the real scientific finding (which is always mathematical in nature) should not be confused as being a terribly good description of reality: plain language just isn't suited to the task.

 

[DevilsAvocado]

... Personally, I've found that things like google, wiki and mathematica are extremely research tools because they make it easier to teach mathematical concepts. A lot of the problem with physics is that you don't have the ability to "play" with objects, and if you have a computer program that let's you play with 4 dimensional hyperbolic spaces the same way that 3-year olds play with blocks, it's much easier to develop an intuition about how things work.
...

(Pure speculations from another layman)

Computers going to (must) change everything
I make a prediction that computers are going to play a much much greater role in tomorrow’s science. Google, Wiki and math software is cool, but is going to look like 'kindergarten' to the next generation.

(Of course I do not know for sure, but) Logically the 'volume' of scientific knowledge is growing for every year, and soon or later we will end up in a situation where the new students have to start reading 24h a day at age of 5, to be able to graduate at 35, and by then most of them will have lost their juvenile creativity, significant for many important discoveries in science. The alternative is to become a narrow specialist, with a very limited perspective, or to work in large teams, which can produce more meetings than science.

Naturally, the scientists of tomorrow need good and basic knowledge, to be able to be creative and to ask the right questions. But why spend 10 years on studying complicated math – if the computer can give you the right answer in 10 seconds...?? My favorite, Albert Einstein, was not a mathematical genius. Many of his ideas started as picturing thought experiments, and later he got help with the complicated math (I know 'reality' isn’t that easy these days...). I think you hit the nail:

The other thing is that the skills that creates a great mathematician often makes for a lousy physicist and vice versa.

So how about a brand new CSS? A Computer Science System with advanced AI that understands your questions, which have a database network with the entire collective scientific knowledge of the latest 2000 years? That collects and explains the information you need for the moment, but not only that; it checks in a few seconds if the math behind your idea is valid! Cool, right!?

And the new CSS will of course provide a graphical environment with 4 dimensional building blocks, to play with. In fact you will be able to personally enter the 4D world! Real cool, right!? (Please, don’t ask about the costs... )

To communicate the new science to the public, computer graphics must play a much bigger role than today. Joe the Plumber will not spend 10 years on math, to understand the first 10% of string theory, just to be informed: "Sorry Joe, it was a wild goose... "

To me, the mathematical representation of 4-dimensional Euclidean space means (almost) nothing:

$$\| \mathbf{x} \| = \sqrt{p^{2} + q^{2} + r^{2} + s^{2}}$$

But this catches my interest, and urges me to check out more at Wikipedia etc:

If the gap between the scientists and the public (and scientist in other fields) grows further, I think it will be very negative for the future of science...

My (highly speculative) hope is that mathematics and computer language is 'morphed' into a 'human scientific language', that don’t requires the knowledge and explanations of 'Egyptian hieroglyphs', before one can start calculate. After all, there is only one basic arithmetic binary operation – Addition of natural numbers. Everything else is derived from that operation...

And most important of all – the kids of today use computers as one of their favorite toys, this will change the world, if nothing else!

(And now I will duck and put my boxing gloves on... )

 

[Chalnoth]

Computers going to (must) change everything
I make a prediction that computers are going to play a much much greater role in tomorrow’s science. Google, Wiki and math software is cool, but is going to look like 'kindergarten' to the next generation.

Computers already play a huge part in today's science. A large fraction of science today revolves around using computers to analyze data, using them to produce simulations, as well as just communicating the results to one another.

(Of course I do not know for sure, but) Logically the 'volume' of scientific knowledge is growing for every year, and soon or later we will end up in a situation where the new students have to start reading 24h a day at age of 5, to be able to graduate at 35, and by then most of them will have lost their juvenile creativity, significant for many important discoveries in science. The alternative is to become a narrow specialist, with a very limited perspective, or to work in large teams, which can produce more meetings than science.

Instead of the amount of education required getting extended, scientists become more specialized. First you need a good basis in the scientific field. This basis really doesn't change much through the years. Then you're trained in your specific area, sometimes with field-specific classes, often by apprenticing to a professor. Often professional scientists like to branch out in a couple of different subfields, but it's not reasonable for any professional scientist today to be an expert in all subfields of a single discipline.

Naturally, the scientists of tomorrow need good and basic knowledge, to be able to be creative and to ask the right questions. But why spend 10 years on studying complicated math – if the computer can give you the right answer in 10 seconds...?? My favorite, Albert Einstein, was not a mathematical genius. Many of his ideas started as picturing thought experiments, and later he got help with the complicated math (I know 'reality' isn’t that easy these days...). I think you hit the nail:

Basically, if the mathematics can't be learned easily by a human, a computer isn't going to fare any better. Complex mathematics may take a few years of training, but most of that isn't so much in grasping the complexity as in just developing a mental toolkit used to solve problems. It's these problem-solving skills that are really needed to be a good scientist, and computers are just a tool that can be used by scientists to do this. Unless we develop real AI, computers aren't ever going to be a replacement for human problem-solving skills.

So how about a brand new CSS? A Computer Science System with advanced AI that understands your questions, which have a database network with the entire collective scientific knowledge of the latest 2000 years? That collects and explains the information you need for the moment, but not only that; it checks in a few seconds if the math behind your idea is valid! Cool, right!?

We already have such online databases. The arXiv stores the vast majority of information about physics, for instance. The problem is that it isn't so easy to just trivially query a store to determine whether a certain idea or other is valid. Granted, mathematicians do make use of certain computer programs to determine whether or not a mathematical proof is valid, but this not only requires some training as to how to translate the language of the proof into language the computer can understand; it is also not something that will actually help with a lot of the mathematics used in science. The reason for this is simple: often the problem with bad math in science isn't that the math isn't internally consistent (which is often a problem), but instead is a problem with the mathematics being incorrectly applied to the physical system in question. And that's not something a computer can test easily.

If the gap between the scientists and the public (and scientist in other fields) grows further, I think it will be very negative for the future of science...

Certainly public communication of science is extremely important, as well as communication between fields, but really, these lines of communication are quite open, as near as I can tell, if only you are willing to look for them.

My (highly speculative) hope is that mathematics and computer language is 'morphed' into a 'human scientific language', that don’t requires the knowledge and explanations of 'Egyptian hieroglyphs', before one can start calculate. After all, there is only one basic arithmetic binary operation – Addition of natural numbers. Everything else is derived from that operation...

That's not reasonable because plain language isn't specific enough. Granted, some programming languages are more readable than others, but human language will never work.

 

[DevilsAvocado]

Computers already play a huge part in today's science. A large fraction of science today revolves around using computers to analyze data, using them to produce simulations, as well as just communicating the results to one another.

That’s very true. I was thinking more of the next generation of IT, which will probably make today’s technology look like an 'abacus'...

Instead of the amount of education required getting extended, scientists become more specialized. First you need a good basis in the scientific field. This basis really doesn't change much through the years. Then you're trained in your specific area, sometimes with field-specific classes, often by apprenticing to a professor. Often professional scientists like to branch out in a couple of different subfields, but it's not reasonable for any professional scientist today to be an expert in all subfields of a single discipline.

But isn’t that a 'problem' that will grow for every generation...? As the volume of scientific knowledge is growing – your 'personal field' is shrinking... or?

Basically, if the mathematics can't be learned easily by a human, a computer isn't going to fare any better. Complex mathematics may take a few years of training, but most of that isn't so much in grasping the complexity as in just developing a mental toolkit used to solve problems. It's these problem-solving skills that are really needed to be a good scientist, and computers are just a tool that can be used by scientists to do this. Unless we develop real AI, computers aren't ever going to be a replacement for human problem-solving skills.

I agree. We need real AI systems before any real change can take place. Still, it will always be the scientist who provides the creativity in the interaction with computers (I think... at least for 1000 years or so...). But it might not be that far away, Chess computers beat the best players in the world now and then, and Chess is all about math. The real 'messy' thing, I think, is to answer 'simple' questions (in the interaction) like; "Good morning! How are you today?", and the 'bright' box answers; "Please define to-day."

We already have such online databases. The arXiv stores the vast majority of information about physics, for instance. The problem is that it isn't so easy to just trivially query a store to determine whether a certain idea or other is valid. Granted, mathematicians do make use of certain computer programs to determine whether or not a mathematical proof is valid, but this not only requires some training as to how to translate the language of the proof into language the computer can understand; it is also not something that will actually help with a lot of the mathematics used in science. The reason for this is simple: often the problem with bad math in science isn't that the math isn't internally consistent (which is often a problem), but instead is a problem with the mathematics being incorrectly applied to the physical system in question. And that's not something a computer can test easily.

arXiv is absolutely great by today standards. Just imagine if you showed this for a scientist in 1950? All papers searchable and available in seconds! Anywhere in the world! What I’m thinking of is the next step in this technology. Today, if you search for "Quantum fluctuation" you get >1000 results, with the first 1000 displayed. All this is based on SQL (Structured Query Language) from 1974, and we need better stuff. A technology that 'understands' the content, not only indexing words...

Certainly public communication of science is extremely important, as well as communication between fields, but really, these lines of communication are quite open, as near as I can tell, if only you are willing to look for them.

Agree, the key is "if only you are willing to look for them". I think many have prejudices about science... difficult... boring... etc.

That's not reasonable because plain language isn't specific enough. Granted, some programming languages are more readable than others, but human language will never work.

Of course it has to be more a programming language than a human language. But is this really impossible? Extremely simplified:

Human:
Take one green banana and one yellow banana, and put them in the basket. How many bananas are in the basket now? There are two bananas in the basket.

Computer:
GreenBanana = 1
YellowBanana = 1
Basket = 0
Basket = GreenBanana + YellowBanana

Scientist:
1 + 1 = 2

 

[Chalnoth]

Agree, the key is "if only you are willing to look for them". I think many have prejudices about science... difficult... boring... etc.

A good fraction of this is down to encouraging better science education, I think.

Of course it has to be more a programming language than a human language. But is this really impossible? Extremely simplified:

Human:
Take one green banana and one yellow banana, and put them in the basket. How many bananas are in the basket now? There are two bananas in the basket.

Computer:
GreenBanana = 1
YellowBanana = 1
Basket = 0
Basket = GreenBanana + YellowBanana

Scientist:
1 + 1 = 2

Try doing it with a more difficult question such as:

What is the comoving angular diameter distance to the surface of last scattering (z = 1089) assuming unperturbed FRW and the WMAP 5-year best fit parameters?

You can state in words what that means, but you couldn't convey how to actually perform the calculation without presenting the mathematics. The most readable computer program would generally just be a simple representation of the mathematics (in this case, it would have to be one that understands how to compute an integral).

 

[DevilsAvocado]

Try doing it with a more difficult question such as:

What is the comoving angular diameter distance to the surface of last scattering (z = 1089) assuming unperturbed FRW and the WMAP 5-year best fit parameters?

:rofl: HAHA! Oh man oh man! I’m laughing off my chair! LOL! :rofl:
This for sure proves who the real nut in this forum is... Intellectual seppuku! Hahahoho! Green banana, yellow banana, I’m going bananas!

You can state in words what that means, but you couldn't convey how to actually perform the calculation without presenting the mathematics. The most readable computer program would generally just be a simple representation of the mathematics (in this case, it would have to be one that understands how to compute an integral).

Absolutely correct! And the right thing for me now is to: "Shut up and calculate!"

Nevertheless, this kind of software must be more effective than a slide-rule. Maybe there is a way to improve this software development? (Providing the user has the right skills from the start.)

 

[twofish-quant]

Naturally, the scientists of tomorrow need good and basic knowledge, to be able to be creative and to ask the right questions. But why spend 10 years on studying complicated math – if the computer can give you the right answer in 10 seconds...??

Because answers are easy, it's phrasing the question that is hard. It takes months often years to write a piece of software that asks a non-trivial scientific question. Also, trying to get the problem in a form that a computer can deal with it takes *more* math, not less.

A Computer Science System with advanced AI that understands your questions, which have a database network with the entire collective scientific knowledge of the latest 2000 years? That collects and explains the information you need for the moment, but not only that; it checks in a few seconds if the math behind your idea is valid! Cool, right!?

Great! Who is going to program this?

My (highly speculative) hope is that mathematics and computer language is 'morphed' into a 'human scientific language', that don’t requires the knowledge and explanations of 'Egyptian hieroglyphs', before one can start calculate.

The usefulness of computer visualization is that it let's you associate a set of symbols with pictures. Once you can touch and feel complex surfaces, then you can better have a sense of what the symbols mean.

 

[twofish-quant]

That’s very true. I was thinking more of the next generation of IT, which will probably make today’s technology look like an 'abacus'...

Maybe. But you do have to realize that a lot of the basic ideas of computer programming date from the 1930's.

Chess computers beat the best players in the world now and then, and Chess is all about math.

It's really not. It turns out that chess computers can beat the best players because chess is a problem that works well if you just do massive numbers of simulations. There is very, very little intelligence in chess playing programs.

arXiv is absolutely great by today standards. Just imagine if you showed this for a scientist in 1950? All papers searchable and available in seconds! Anywhere in the world! What I’m thinking of is the next step in this technology.

The next step is really to get people to summarize those papers and find the three or four that are worth reading.

A technology that 'understands' the content, not only indexing words...

Or a technology that makes it easy for humans to comment on the content. If I hand yoiu a thousand papers on string theory, it's going to be totally useless to you. What you need is someone to go into those papers, add commentary, filter out the dross, put together summaries, cross reference them to other sources. We can call them teachers.

Of course it has to be more a programming language than a human language. But is this really impossible?

I don't see the point in getting a computer to do what a human is good at doing. If you want to query a topic, instead of an automated search, it would be better to have a place where you can type in a question like "I'm looking for a good introduction to deRham co-homologies, where can I find one" and then have some group of human beings answer the question.

 

[Chalnoth]

The point is that computers are merely tools. Extremely useful and capable tools, this is true. But they cannot replace genuine understanding of the mathematics and concepts being used.

 

[DevilsAvocado]

... computers are merely tools ...

Agree, but if Max Tegmark is right, it’s going to be quite powerful tools in the far future, since you then will be able to simulate complete worlds...

 

[Chalnoth]

Agree, but if Max Tegmark is right, it’s going to be quite powerful tools in the far future, since you then will be able to simulate complete worlds...

I don't think so. The number of degrees of freedom for an entire universe is just too vast.

 

[DevilsAvocado]

Because answers are easy, it's phrasing the question that is hard. It takes months often years to write a piece of software that asks a non-trivial scientific question. Also, trying to get the problem in a form that a computer can deal with it takes *more* math, not less.

That’s true. Just my wild speculation: If we solve the IA problem in the future – what if it then would be possible to construct a 'mathematical internet' that is 'aware' of what has been tried out, what works, and what’s going in on at moment? Just as you today go to Google or Wiki and type word or a phrase to get information – this system would 'assist' you in the mathematical world...

Again, it's wild speculations... but of if you don’t have any visions about the future... probably less happens in the future...

Great! Who is going to program this?

Maybe I can put a team together... can you provide the bucks...?

The usefulness of computer visualization is that it let's you associate a set of symbols with pictures. Once you can touch and feel complex surfaces, then you can better have a sense of what the symbols mean.

Exactly! Let us move one level above pen & paper, and make things that move in 3D/4D, to move in 3D/4D!

 

[DevilsAvocado]

Maybe. But you do have to realize that a lot of the basic ideas of computer programming date from the 1930's.

Or, maybe 200 BC? When an Indian writer presented the first known description of a binary numeral system!?

It’s all about 1 and 0, but the thing that changes is ability to process data and calculate faster and faster. Also the tools for developing software have gone from punch card to graphical RAD (Rapid Application Development), which speeds up everything by many factors.

Personally I think we need better/smarter software to make a leap into the future...

It's really not. It turns out that chess computers can beat the best players because chess is a problem that works well if you just do massive numbers of simulations. There is very, very little intelligence in chess playing programs.

Agree, basically what the program does (when leaving the opening book database) is testing every possible move (while the opponent speculate what’s for dinner ), save the evaluation, and then select the best move.

Maybe there is no possibility to translate real intelligence to 1/0... Maybe one way is to find out what real intelligence is, and how it 'works'... first...

The next step is really to get people to summarize those papers and find the three or four that are worth reading.

Yes, another thing one could do tomorrow is to extend the arguments for SQL:
* How many have required the paper (last month, year, etc)?
* What’s the score (usefulness)?
* What’s the average author score (usefulness)?
* How many references to this paper on internet?
etc ...

Or a technology that makes it easy for humans to comment on the content. If I hand yoiu a thousand papers on string theory, it's going to be totally useless to you. What you need is someone to go into those papers, add commentary, filter out the dross, put together summaries, cross reference them to other sources. We can call them teachers.

Comments yes, and 'scores' etc.

I don't see the point in getting a computer to do what a human is good at doing. If you want to query a topic, instead of an automated search, it would be better to have a place where you can type in a question like "I'm looking for a good introduction to deRham co-homologies, where can I find one" and then have some group of human beings answer the question.

Human interaction is always good. But if you apply this reasoning on Google it becomes a 'little' hard to justify... "I don’t want any stupid Google! I want to call my professor in the middle of the night so he can recommend some really cool webpage to me! God damnit!"

 

[DevilsAvocado]

I don't think so. The number of degrees of freedom for an entire universe is just too vast.

Okay, the smell of science fiction is overwhelming. But on this I’m 100% sure Dmitry67 has the opposite view...

 

[Bodicea]

Who is "we"? Most physicists don't have any problem intuitively grasping fourth dimensional spaces.

ok then, explain what a 4 dimensional environment looks like...assuming of course that the 4that you are saying that the 4th dimension is a physical manifestation...cos otherwise you are talking about us living in a 3D world with 'time' being the 4th dimension..and a 5 year can intuitively 'grasp' that. :)

 

[DevilsAvocado]

... explain what a 4 dimensional environment looks like ...

How about this graphical representation of 4-dimensional Euclidean space:

 

[Chalnoth]

How about this graphical representation of 4-dimensional Euclidean space:

[PLAIN]http://upload.wikimedia.org/wikipedia/commons/5/55/8-cell-simple.gif[/QUOTE][/URL]
Well, that requires a bit of explanation. That is a rotating 4-dimensional hypercube projected onto three dimensions, and then re-projected onto two dimensions for visualization on a computer.

A hypercube is the generalization of a cube, which is, in turn, the generalization of a square.

A square is a closed object in flat two-dimensional space with sides of equal length connected by right angles.

A cube is a closed object in flat three-dimensional space with edges of equal length connected by right angles in each direction.

A hypercube does the same thing yet again. Just like in the transition from two dimensions to three dimensions, there is now a new right angle available. So instead of three sides connecting to each vertex, you get four sides, all at right angles to one another. And just as how with a cube the sides are squares, for a hypercube the 'sides' are cubes.

When you see the hypercube projected into three dimensions, as above, you see the various 'sides' (which are cubes) projected at different angles: they look like cubes distorted so that their vertices are all connected. But what's really happening (mathematically, at least) is that the hypercube is just rotating in 4 dimensions, and the projection becomes distorted by the fact that we can't represent 4 dimensions in 3D.

In the end, I'm not sure it really helps to give us an intuitive understanding of four dimensions, but at least it looks cool.

 

[Bodicea]

How about this graphical representation of 4-dimensional Euclidean space:QUOTE]

Devils advocate...the 4th dimension is time, and time is a concept, not a physical manifestation. What you have shown us is three dimensions within three dimensions. That CGI is dispirit, and there is no continuity that shows the 4th as being a physical manifestation or even a component of the 3D world. That diagram is further evidence that physicists are really grasping at straws to proof their theories...It is simply a cube within a cube

Also, as someone living in a 3 dimensional world it would be impossible for me to represent the 4th dimension physically. Its like a 2 dimensional entity trying to show others in that dimension what a 3 dimensional world looks like. They cant,cos even though the third dimension is there, just at their fingertips, they can't represent it physically cos they live in a 2 dimensional world and the third dimension, to them, is like the 4th dimension to us...a concept and something we cannot actually produce.

And why do physicists need more than three dimensions to move forward, when all the answers we have achieved are explainable within the constraints of a 3D world and all the most recent findings/discoveries are confined to the three dimensional world, from quarks to the cosmological expansion. The only reason multiple dimensions are needed (imho) is to facilitate the parrallel universe thesis, and I consider that a load of fertiliser

 

[Chalnoth]

Sorry, Bodicea, you're way off base here.

First, extra dimensions have nothing whatsoever to do with most any multiverse hypothesis. They're completely and utterly different ideas.

The possibility of their existence is an interesting one, and is being actively investigated on a number of fronts. One reason why people are looking into the possibility is that if we propose that the fundamental objects are strings instead of particles, we are forced to conclude that there must be a total of 10 dimensions of space. While string theory is very, very far from being experimentally tested (or testable), some aspects of it are quite testable, and the potential existence of extra dimensions is one of them. So this is one of the things we are looking for when we examine gravity's behavior at very short distances, and the behavior of subatomic particles at very high energies.

No physicist today would tell you that more dimensions are needed to move forward, but string theory has a number of very exciting features to it that make many think it is a promising (but extremely difficult!) theory that may potentially explain our universe. And the nice thing is that experimentally, searching for extra dimensions experimentally increases our knowledge of the way our universe works regardless of whether or not they exist.

Finally, let me point out that we know that our current theories of how the world works are insufficient: they are inconsistent with one another and have internal problems that ensure that they cannot be completely correct, and instead can only be approximations to the true behavior. So, when we do things like increase our volume of cosmological observations or build instruments like the LHC, we are looking for new experimental regions where our theories no longer give the right answer, so that we can have some clues as to where to go next. Today we have many ideas, and they may well all be completely wrong. But that's fine as long as nobody is claiming that these ideas are true before they are experimentally verified. And nobody is doing that (except for a few cranks, I suppose...).

Oh, and one final point: relativity forces us to conclude that time is every bit as real a dimension as the three dimensions of space.

 

[Skolon]

Or a technology that makes it easy for humans to comment on the content. If I hand yoiu a thousand papers on string theory, it's going to be totally useless to you. What you need is someone to go into those papers, add commentary, filter out the dross, put together summaries, cross reference them to other sources. We can call them teachers.

This remember me about the huge knowledge base from Trantor (from Asimov's Foundation series). As I remember, an automated search system like yours is described there. Who knows, maybe we will be able to create something like this not so far in the future.

 

[Skolon]

Or a technology that makes it easy for humans to comment on the content. If I hand yoiu a thousand papers on string theory, it's going to be totally useless to you. What you need is someone to go into those papers, add commentary, filter out the dross, put together summaries, cross reference them to other sources. We can call them teachers.

This remember me about the huge knowledge base from Trantor (from Asimov's Foundation series). As I remember, an automated search system like yours is described there. Who knows, maybe we will be able to create something like this not so far in the future.