Every once in a while, reading articles at CleanTechnica, I feel an urge to overcome my reluctance to put myself into the spotlight by writing about my own inventions. I am afraid that Steve Hanley, who happens to be one of my favorite CleanTechnica authors, put up the straw that “broke the camel’s back” when he wrote “None of us have ever invented anything” in the article “Fraud, he wrote! Whistleblower Takes On Donut Lab Battery Claims.” The truth is, I invent things from time to time. And considering my closest relatives, the majority have patents. Reading what Steve wrote, I decided it is time for me to comment, because I can see that we, in our society, have to do better than we are doing, both nationally and worldwide. I want to comment on this using three inventions, two of which are mine and one of which was produced by my brother, Reid. Two of these three inventions were patented, though one of mine was not. Each represents one of the reasons why potentially useful inventions are never brought to market. Please note that I am not going into the most common reasons why inventions are not patented or do not earn inventors any money. My father, for example, invented procedures relating to navigation while he was a Navy officer. That was a funny story for another time — though, this meant the invention belonged to the Navy. And many employers require intellectual property agreements stating that they own inventions of their employees, regardless of whether they relate to work. I am not dealing with these here. Invention 1 A little over fifty years ago, I woke up with what felt like a complete understanding of a modular geometric shape. It is funny how such a thing can happen, but it did. What had come into my mind was a modular construction consisting of a floor made up of equilateral triangles, each subtended by a single isosceles triangle. I call the shape a polymodular, because it is made up of repeated use of a set of four modules. (Skip this paragraph unless you want some technical stuff.) The smaller of the isosceles triangles is such that six could be the sides of a regular hexagonal pyramid, with each side of the base 2 units long and the altitude of the pyramid at 1 unit. The larger could be the side of such a pyramid with a base 2 units long and the pyramid’s altitude is 3 units. Interestingly, the triangular sides of the pyramid built of the smaller triangles would slope at an angle of precisely 30°, and the triangular sides of the taller pyramid would slope at 60°. Aside from these two triangles, only two others would be needed. One is an equilateral triangle with sides of 2 units, used for a floor. The other is a right triangle with a base of 2 units and an altitude of 3 units, which would provide vertical walls so the modular structures to be fully enclosed. The advantages: Shapes of such a type could be shipped in little space and used to build emergency housing very rapidly. The housing could be changed quickly, so that while buildings could be small initially, they could also be quickly built on and extended. They would make for a very strong building, and it would be very inexpensive. And they could be built to be very comfortable. The problem: I sent this design to the inventions management division of A.D. Little. After I waited for six months for a decision, I called. When the operator learned who I was, she put me straight through to the president of the division, and I was talking to him within seconds. He knew all about my invention. He said the delay happened because the lawyers could not agree with the engineers over what to do with it. The engineers said it was potentially important, but the lawyers pointed out that it constituted a geometric shape, and geometric shapes can’t be patented. Conclusion: Capitalists must be paid sufficiently to earn money, or they won’t invest. In order to make money, an invention typically has to be patented. Without that, investors will not put money into it. Invention 2 About forty years back, I hit upon an idea. I was doing computer programming at the time, and found the many designs for variable storage formats really interesting. Of course, they were almost all for storing numbers, with the only exceptions being those that stored alpha-numeric characters. I asked myself, “What if there were a way to store such non-numerical information as words and phrases?” When I told people what I was working on, they nearly all told me I was crazy. I kept on going, however, and after twenty years or so I had an initial design for a format I liked. (Skip this paragraph unless you want to read some technical stuff.) It is conceptually somewhat easy for some people to follow. All language is divided into eight types of concepts, which are then divided into four subtypes, and these are then divided again into eight types. For example, the overall type “agent” includes such things as living organisms, machines, and so on – anything that can do something. One of Agent’s subtypes includes “mechanical,” and one of the mechanical subtypes is “vehicle.” Altogether, there are 256 concepts in the base vocabulary. The vocabulary is divided made up of 64-bit words, each with eight fields, and each field is represented by the 256 word base vocabulary. Each field has a specific meaning, such as “source,” “purpose,” and “context.” Finally, each bit in the 64-bit word meaning is marked as to relevance in a 64-bit mask. Thus, each word requires 128 bits for full representation. The advantages: Basic representation of a word or phrase only requires 128 bits, and these can be calculated in a normal CPU using normal operations, just as numbers are, though an ideal CPU would include use of the filter. This is all pretty obvious CPU design. It means that if you want to find the difference between a taxi and a pedal rickshaw, the operation could be done in a single clock cycle, assuming the registers are pre-loaded. Then, integrating that difference into another, preloaded word takes one more clock cycle. For example, integrating that result into “motorboat” will produce “pedal-powered boat.” If the job cannot be done (such as integrating the result into “wax”) the result will produce a fault. The problem: Whether people just didn’t have time to take a good look at this, or whether it was a matter of it being “not invented here,” it turned out nobody really even took time to look at this invention. This is despite the fact that it received a patent and was protected. There may have been as many reasons for not looking at it as there were people who were asked to look. My own thought on this is that it represents an approach to AI that could require a tiny fraction of the computer power that is driving tech companies to think about building nuclear reactors. I realize that this approach cannot do such things as image processing, but I think it is too powerful to ignore. For one of my tests, I had a computer take two inputs, find the difference, and then find every pair of words in its database, at that time about 2,500 words, with the exact same difference. Inputs of “car” and “small passenger airplane” produced such combinations as “commuter bus” and “commuter airplane,” which were each one word. The same test produced one word combination that I really had to examine. It said, in effect, “As a car is to a small passenger airplane, so is a lion to an eagle.” I was perplexed. Why lion an eagle, and not lion and owl, or bobcat and falcon? It turned out that in 64 bits, the definitions of lion and eagle both said they were symbols of authority, making the match exact. Conclusion: If no one looks at it, no one will invest. An invention has to feel non-threatening and have something attractive about it. Invention 3 This brings us to the third invention, which was produced by my brother, Reid. He invented a special ceramic filter. It was made to allow water to pass through slowly because it is slightly porous. He had found that the addition of a tiny amount of silver to the clay would make it toxic to viruses and bacteria. He developed a couple of ways to apply the silver. And he got patents. The U.N. once tested several filters to determine which was best for removing bacteria. Reid’s was rated as the best. The government of Bangladesh tested it by putting water through that had been taken directly from the Ganges River. The bacteria count was zero. (!) I have to admit that Reid’s filter does not remove toxic chemicals, such as arsenic compounds, which are a problem for Bangladesh. But it seems to be a big step toward safe water for many people who don’t have it today. And the advantages are obvious. The problem: Of the three inventions presented here, Reid’s might be the most valuable, with the greatest possible good coming for the greatest number of people. The biggest problem with Reid’s filter can be that it can be made by just about any village potter, anywhere, to be sold for about $5. And because it can be back-flushed to be cleaned, the filter will last until it is broken. These may sound like positive things, but they mean that big business can’t make money on Reid’s filter. It happens that there is one factory making these filters. It is in Nepal. Maybe people in Nepal are more interested in keeping people healthy than they are in making money. Conclusion There is a broad problem. In order for capitalists to invest, they have to see the possibility for a return on investment that is large enough to be attractive. It has to be big enough to make it worth while risking a loss. Unfortunately, there is also the problem that without the greed that drives capitalists, people are not likely to put enough time into investigating a new idea to understand it fully. And without that understanding it is easy to walk away from a valuable, misunderstood invention. This implies that socialism is not a cure-all. I think another type of social evaluation might be necessary. It could be one based on the potential good versus potential cost. That may require invention of a new approach to investment, but I feel certain it is cannot be based on greed.