Wednesday, October 26, 2011

Are Cybersecurity Problems a Mirage?

Remember word processors? Memos are great! Scott Borg in what I think is an amazing article because of a simple omission he may seems unaware of says malware is what he calls a "cybersecurity nightmare." To quote,

"The world of cybersecurity is is starting to resemble a spy thriller. Shadowy figures plant 'malicious software' on our computers. They slip it into emails. They transmit it over the Internet. They infect us with it through corrupted Web sites. They plant it in other programs. they design it to migrate from device to device-laptops, flash drives, smartphones, servers, iPods, gaming consoles, copy machines-until it's inside our critical systems. As even the most isolated systems periodically need new instructions, new data, or some kind of maintanence, any system can be infected."

He then goes on to list sabotage problems like planes crashing, trains colliding, health machines failing, banks being discredited, poison drinking water and so on. Then he continues with what we can know about malware to stop it, e.g. we can't read the code right before our ears with my vision how, right?

We can't find it in hiding because of many places the machine can't read the code, the code to check the bad code may already be tainted, and so on.. Sharing is alright if health not illness. Hope you recieve a card from Blue Cross, says Hope You Feel Rich By This Month! If you go to the doctor and she Says, the Opthamologist will see you now, your eyes are brown and you see right...

My simple question about malware is;

Why weren't there the same problems in the old days, say with the security of the banks or the power wires, my old word processor or when mom calls? (Thanks mom each Mom's day each time you call me..)

The answer could be in simplicity only up to the point where you couldn't know if you win or lose, (If you don't know how you know what you know, you lose more than otherwise"). Consider the three ways computers send code in the second paragraph; instructions, data, or some sort of maintenance. The "wiser" machines of "older tymes" had much more of the data sent, not "instructions or maintenance". People did all the machine operation and maintenance with stuff like business agreements and hand shakes. The wires were secure, the copy machines were secure because there wasn't all the extra add on nonlinear stuff between the machines that can't be known so well. I always thought it was so idiot about the slow upgrades I would see taking up hours of my time, by surprise. Why do I need these, to go in the library and read, or to send a letter was simple and no problem in the old days. This is why I say the middle of the three, of instructions, data and maintenance is too complex. My solution would be to stop the codes for data by converting it to analog first and sending it high speed, then the machine reads it much like an older letter by an onsite sensor, ect. Edison imagined the wires of his Ma Bell machine to be used to send automated letters. A Letter is a letter and if you ask me and doesn't have to be a number so much. This method using microfilm to store the web sites once edited would simplify the process of sending documents like memos, all one simple type of machine. And it's secure. This would solve the problem of tainted malware web pages. There are mechanical computers that are much faster and more energy saving by far than electronic machines. I was a bit too aware like mom the journalist when we found we couldn't buy a simple word processor anymore in the old days. It was simple without all the idiot complications of a PC.

Once data was more sound, like a great library, the question of instructions or maintenance might be solved by just doing without. We didn't need them then or now or we certainly need a way to adapt to the need to change so the foundation is sounder. I was a musician for years to no avail, I tried many ways to remember and be good. Finally 25 years later I got wise 10 years ago then and realized slower but surer is more the merrier, evolution takes time, and knowing what you get is also how the chess machines win. If you make a move where it understands the payoff it will chose to make a move where it understands the payoff over moves where it doesn't know, this is wiser and how it would win. As I listen to my better older songs I found that if I would improvise so for any one mistake I did it three times the the more sound way, even if I was unaware of exactly what I got, it was more and more evolving, a sort of Bayesian method used by many computers. Loose lips sink ships, so even if this makes a great song using both the method of improvement by question and answer "like science" for the song and then memorizing by saying the words of what I wanted to remember in each loop of the rep won by the science method to set boundaries and structure on my improvisations won by the three out of four method also, this method is not the same as for something like a cheap war machine or where a person's life is at risk if harm was so.

Because slower but surer is more like the old days. What about the "modem wars" you may ask? The modem wars are where the army is being attacked millions of times, faster than humans can think via modems. If the US stopped all the codes by the above it would take a while to convert all the machines to a complete system without code between any of them. This is a step that would slow our machines a bit but the malware problem would be solved. As we are vulnerable due top the web, any nation that's without a secure system is too. Thus while it would take awhile to convert out the code we would be more able in war not less if we do. If other nations didn't also convert, we might lose some business, and it would cost more for the machines. I think this is like the middle ages, if they had a castle and it was secure, though not a complete luxury like the web often is here, it would make it so we could make it to the more advanced methods if we realize here we may pay a lot more if we allow the codes than if we don't. This would involve much the same method as ye old word processor, careful checks by simple more reliable machines we know we can trust, no demons or dungeons, the older ways were best. Decartes would say in the old days, I think therefore I am, (they spelled wrong on his 400 year commemorative stamp in the Old World www.) The idea here is simplicity of our machines is predictable, but only up to a certain level, if the machines are always below this level of complexity, we would win more than without. This also relates to the problem of the complexity of the business of making chips has gotten so international bad characters can sometimes put hardware in that may cause problems months or years later, or why not not at all? If all our machines are with our own ways to solve like the royal days, and the code can't be sent, all shall be well, hope so.


The hardware problem could be solved by standardizing the the chips by one common type of chip so there's no hardware where bad code could hide. I think a good solution might be the IP of molecular crossbar chips. These are a simple mesh of wires with a bit where each wire meets. This is a real simple chip to make. Some might object to the lower speeds this type of chip has even if cheap to make. A solution here that's been invented might be to perhaps use a particle as a bit where the wire meets and as it rotates around in mostly continuous space it would store the angle well. Each small bit has three components; a write component, the particle that moves at libitum, a brake on the particle by another to store the angle, and a read component that only would read the brake, which itself doesn't much give way when being read. Since space unlike a quanta is continuous, and the field that is stored or changed is much more precise than a quantum, even one of these by storing the angle could store perhaps an 80,000 bit number. And it's nonlinear so even two of these small bits could store 80,000x 80,000 times as much as a base two chip. This would be super low power in consumption and high in efficiency, and super fast. So the standard chip would be of value even if simple.

About the code, there three levels, the top level is the general method of using prime numbers to agree to the transaction. This level is probably secure because it's based on the well known idea that it takes much more number crunching and division to factor out a code than it does to multiply up two numbers and create the code, the standard method of cryptography currently used. It's based on the ancient science of number theory, and since it involves both division and multiplication, it also involves addition and subtraction, and new operations other than these four operations are not likely to be found. If it were not mostly sound the web would have long since crashed.

The third level is just the binary ones and zeroes the machines encode from the prime numbers for ease of making simple chips.

The second level, however is where the problem seems to be after all a number is just a number without giving instructions to a machine. This is why even a single number out of place can down a machine. The machine can't understand if it's under attack, it's mostly a robot per se.

At the dawn of the internet we would hear of the wonders of all computers speaking the same language. Actually this isn't a language yet, rather just a series of instructions, leading to my thought about the code; how is it we understand when we speak and we know the other person isn't trying to hit us (booses are the exception, hint boss don't read this). The answer is that we share a language with rules of grammar and vocabulary, ect. This isn't just code we always accept without hopes for more. To say computers always speak a common language seems like saying a copy machine is reading and understands the volume. And here may be an important solution or improvement to the problem of bad code.

If we reedit the code and make it like a real language with steps like we speak in sentences, in two part components, noun verb found in all languages, with rules of grammar for each step, the machine could perhaps know what's allowed or to be stopped. In essence, the noun verb would always be about physical power just as in brain research any change in ideas has a change in physiology in parallel, ect. Another form of the noun verb might be a question and answer sort of like Plato's method of dialogs. (Why are you doing this? Like a claim and proof in rhetoric, a thought the straight line and the punch line in comedy, a setup a bluff and a move in the NCAA, or the idea of Noether's that all there is can be classified by what changes and what stays the same, important to physics and the foundations of math.). If the computer doesn't agree to the claim it could stop just that claim and save the rest, leading to much reduction of computer idiocy.

My mom the English professor says, speak and say well, real well, not real good. You look up the definitions of usage and, like who cares, there is no standard usage just a hope of marms to control the language that never has won or would hope to win. With computers on the other hand her utopia might be seen! The machines might not fail if all definitions are set at the outset. This is essentially what we didn't achieve when the codes of the second level were founded. Like the synapses being much more in area than just neurons like other life, making the code a real language might be both much more reliable and economical about efficiency than no codes as in the above. The only problem here may also about standard use since presumably there might be more worth in evolving the language just as with real language, e.g. because of per se economy to make cheap machines. Even so I believe redefinition of the web by this method may save it from malware.


I've defined the three levels of the code, as with the "simple language".the possibility of higher levels of the code; this would be like a spy who sends a letter to another with code words or letters embedded in the usual sentence. However if the letter is only read at the middle level and no higher because the machine reading the letter by what it knows is well proven, no bad code would be involved. This would be like a third person like me reading the spy letter who would, say, Well Duh! No doubt the machine might call the boss so the higher code is stopped, even so real language might well stop malware.
..