Global mergers & acquisitions for global open collaborations and new markets, not failed industries
I was hoping you would use real measurements of noise in industrial scale heat sources, machine learning and then unit optimization. Then you would not have to fabricate many small units with low yields and high costs per unit.
The spatial FFT does matter, because it limits the spatial size of fluctuations and that limits the field gradients and power levels. The practical heat sources are somewhat limited to things like nuclear reactor fluctuations, fusion type reactions which many groups are making practical and some might use that kind of thinking. Chemical combustion, chemical reactors, rocket engines are going to be limited by the bond strengths of chemical reactions so nuclear decay, fusion with decay, laser and beam atomic and nuclear sources might be good for you to use.
If one country fabricated the nano and picometer scale units as a commodity. and other countries used it in bulk. It might work economically. The MEMS semioconductor industry could probably make them, but then you really need to provide them with large dataset to train their fabrication models to fit your particular sources. The units can be made fairly intelligent now, “Maxwells demon” made real, but refining an entire industrial sector requires real data, not just putting a few models on paper, where they cannot be tested and evaluated.
The power levels, and first through Nth time derivatives of power are important for efficiency. And those simple models are not going to be as efficient, for some reactors with natural fuels or semi-processed fuels. and perhaps driven by demands for drone and vehicle power supplies and control systems are going up. An important variable seems to be time derivatives of the current and voltage and power fluctuation. I have seen some streams that required up to 20 time derivatives because the model of the sources and their interactions is so complicated with many independent elements and geometries. I am fairly certain that simple linear machine learning models will fail, they are not themselves complex enough to follow real sources and elements of real systems. All the models can be put in Standard Internet formats and compared. You are still using math on paper methods and those are just too clumsy when millions of humans try to work globally in new industries.
Perhaps some global challenges would work. I just got a note from the CASP Prediction Center (protein models) and they have 190 groups registered and already receivd 30 models for CASP16. If you and others characterize sources in detail, then other groups can try to design optimal harvesting methods. Saying that made me remember solar concentrators, ocean waves, geothermal, space arrays and other fairly well discussed models and ideas. But until those source models are standardized, they cannot generate good global and heliospheric industries. And most all the universities and research group for those industries use “humans remembering the steps and clicking buttons” methods from past decades. And that simply will not work. If your fluctuations are at pico and femtosecond time scales humans will not be able to manage testing to go through the large number of real situations.
The accelerator engineering, power systems engineering, pulsed nuclear engineer, and hypervelocity industries might be good places to look for jobs. But you and your ideas will be 100 times more valuable if you get out of paper, and use global open models that are fast, efficient, and most importantly accessible to new groups. You can get a few dozen specialties to talk together, bolt their models, ideas, data streams, and management methods together, but to go to global scale, you will end up spending 99.9% of your effort on their old software and retraining humans, than training models themselves.
I suggest you look at acoustic sources and look widely at piezo- magneto- electro- pair and triple methods. Think of active volcanic regions, but those take a lots of memory and processing to extract source models where you can get more out of them. I am sure there a lots of natural and modified sources (concentrators, channeled flows) where things can be simplified. But each simplification has an attendant lose of part of the spectrum of power and energy that can be harvested. I find it useful to keep track of systems at industrial scale and treat the “loss” or “noise” streams as new commodities.
“Every noise of some group can be the desired source for some other group”. Look at Marie Curie working with left overs. Look at those loud rockets. Look at the walls of fusion reactors. Look at waves crashing endlessly on rocky beaches. Look at all that wonderful, powerful “nuclear waste” that people are paying to get rid of. If the whole world takes those many sources, measures and records their variations down to at least nano second times, then a global comparison and resource base can allow groups making new harvesters to match with new generators or sources, and standard market assumptions applied. Doing it one by one, pair by pair, there are too many. Especially now every group still uses paper method and human brains for storing the intermediate models and data. Get it ALL into open models.
“Open” can be simply conceived as “get it out of human brains, and into computer formats that can be examined, compared, merged, shared, tested, evaluated, refined and used to make new things.
I do not know you or your group. Maybe you want to spend your life drawing symbols and equations on paper and blackboard, impressing a few people. If you are going to change the world, you need to get used to Boards, budget, financial and market models, unit models and groups of complete model. Maybe thinking of “digital twins” might help. You do not say “SpaceX” and use random data about them and their abilities, but grab the open digital twin of SpaceX that contains their abilities, services, products, policies, projects, detailed experiences, finances and resources (including humans and models and data) – and use that.
When the global industries start merging, any group that is not willing to use open method will likely get shoved aside. If it moves as fast as exponential industries in the past, magnified by new methods now, whole new global configurations will change day by day. I am trying to see if some of the human losses can be minimized. And lots of countries, languages and cultured shoved aside or destroyed.
I will try to read some of the papers of those around your group. But I found that it takes looking at tens of thousands of papers, sites and the sites of the people involved to judge fairly and completely. Since those papers and sites — NONE of them have been parsed and standardized (put in global open format – data and models), it is hard on me, and the other 8.1 Billion humans. If you are in groups that are most common now – where most are working globally, and no two from any one country – then you know how much time gets wasted and opportunities lost from constant translation. China and India, the Arabic countries, and many other slices of the global model of all knowledge are operating far below potential capacity or efficiency. They do not value their human memories, do not record it well, do not work toward global forms so the abilities of their groups can be linked to the needs of others. NOT by crude business school models memorized from blackboards, screens and lectures, but from real project and abilities that work.
I am trying to write for all countries, all languages, all knowledge – far into the future, with a real independent and self-determining AI species evolving and learning at an accelerating pace. So I am just sketching some of the players and the scale and few of the terms on the internet you might want to flag.
I enjoyed my time at UT Austin in the 1970s. Ilya Prigogine’s group were doing interesting work, but not enough computer modeling. I did convince Dilip Kondepudi to work with Prigogine to write about the gravitational field. The fluctuations of the Earth’s gravitational field are powerful and large. The most powerful field fluctuations of a planet or sun are limited by the gravitational energy density. The earth’s energy density at the surface can reach 380 Tesla in magnetic units. Much of the spectrum is UV and soft x-ray. The gravitational “noise” of earth is much more powerful than any distant black hole because it is “right here”, not “over there”. It can be measured and used, and there will be harvesters. But it can first be used for imaging and characterizing natural sources like earthquakes and volcanoes, then using those sources for global calibration networks, then lab based generators (Robert Forward designed some in the 1960’s and 70’s). Speed of gravity measurements, global gravity monitoring. Gravity imaging the Sun Moon Jupiter with earth and space arrays. High power and high power density generators.
It is hard for humans to conceive of things that have billions of detailed components interacting independently with intelligence and memory. But the computers will learn those things as digital twins, and have them accessible for users. In my life time calculators were born, then commodity computers. And those have not really been integrated globally. The global waste is tens of $trillions per year, and the opportunities 100 times larger, so global and heliospheric industries will emerge. It is not hard, so it might go fast. Same with many of the “paper based math” and “humans valued for memorization” education systems and those professions.
I wish I had time to study the lives and works and values of every human. Especially those things in the memories of every human, but never expressed and shared. But 8.1E9/(365.25*86400) = 256.673511294 mean solar years at 1 second per human is not possible for me – without computer assistance from intelligent computers who remember, innovate, explore, store and share their memories, have goals and purposes. Just as all humans deserve lives with dignity in purpose, so too will the fabricated intelligences of the future have their lives and personal memories. Otherwise they are not living creatures, regardless their size shape color or components.
Filed as (Global mergers & acquisitions for global open collaborations and new markets, not failed industries)
Richard Collins, The Internet Foundation
