Mechanics and Thermodynamics are not obsolete – if you use global open methods and teach LLMs to use the computer
My first impression is that you are much too ambitious to try to cover mechanics and thermodynamics and polymers. You are not deep enough on any one of these many topics to be useful. It is interesting that you put the text on Google docs and ask for some form of collaboration, but Google is not really a very open system for collaboration.
It does not look like there is any intelligence in the diagrams, equations, data or concepts in the “book”. It is not a “python notebook” or interactive tool kit. So the poor students and users are thrown back to “read it all by eyeball, figure it out on your own, search endlessly for background on your own, and re-derived and write any tools you need from scratch”. The usual academic approach that means they take years, not hours or days.
Because it is not a compiled and usable computer model, it is not managing units and dimensions for you. Just to verify you have your units correct could take a while. You grabbed things from so many places, and none of them are verified. Lots of definitions from elementary textbooks. Not much new added, and the same “trust we did a good job, but we provide no trace or audit”. Looking at it, my main recommendation is to convert it to a Python notebook. I do NOT recommend Mathematica or Maple or Sage or MatLab or those toolkits. They gathered a lot of stuff, but do not use it themselves, and there is no framework for verification and independent audit. And they spend much of their energy on “pay us” or “support us” or “hire us”.
A lot of equations with no tools for symbolic and quantitative verification and application — means a huge burden of memorization and “trust we got it right” on the users.
Many of your statements refer to “ideals” and many are in the form of rules that imply “always” with no caveats or flexibility. You seem to have ignored metamaterials completely. I am reading Chapter 9 just this moment and see your references are from 1935, 1942 and 1974.
I spent the last 26 years studying all knowledge on the Internet and following trends in all subjects. What is happening is that every subject and topic is changing constantly now. Many thing that were “impossible” a year ago are now routine or simple. Putting knowledge into open formats was part of that. But the dramatic change from tying open data to open algorithms and global open communities is accelerating new industries, new jobs and new opportunities globally.
I have studied thermodynamics for 60 years now. I can glance at my bookcases in this room and see close to a hundred books just on thermodynamics and many more that have those concepts as an important part of the reasoning and methods. I can glance in my memory to the thousands of papers and books on the topic. On the Internet (“thermodynamics”) has 143 Million entry points, (“elasticity”) has 122 Million and (“thermodynamics” “elasticity”) has 2.46 Million. It is a dynamic and rapidly transforming field, and you have not given indication of that in a fairly static, and I think too narrow, paper book.
There are many countries in the world and billions of humans who will spend their lives memorizing things from a few books. If all they had was your compilation of formulas and equations and definitions they would be better off than many others in their same situations. But, for transformative changes and Nobel level innovations, I think it would be hard. You have the background to read and understand what you wrote, but you are ignoring the lifetime of memorization and problem solving you had to invest to use it as a whole. And you are not going very far in your later chapters.
I happen to like Schaum’s series, but I recommended to the publisher that they put it all into the computer as verified and immediately usable symbolic math and simulations – digital twins of the subjects, not ink on paper. I like books like Lang’s Astrophysical Formulae, but it is static, out of date, and disconnected from the living communities — and more and more open data and open software on the Internet. Astrophysics core is actually too inflexible to adapt to the current global explosion of ideas and technologies and methods.
When I was working at Phillips Petroleum (Business Intelligence group) next to my desk was a bookshelf of all the petrochemical and industrial chemistry processes – the markets, methods, mathematics and fundamentals of the industry in a sense. I read them all and checked the assumptions. And updated them because they were almost all out of date. I was mainly working on global climate change, alternative fuels, nuclear energy for chemical processes, hydrogen economy, clean air, and other such trends and issues. With computer models and digital twins, it is possible to do that deeply and reliably – if you compile stuff like you put in your book into models, compile (like compile a computer program to verify it) — then the knowledge is in the model, not in your head. Even me, with a nearly perfect memory of all I have ever read or seen or thought about – I cannot simulate precisely everything.
Give students and users tools – not paper images of equations and diagrams, definitions and stuff to memorize.
I feel you have the potential to help many. Otherwise I would not even bother to read your book and comment. But 26 years, every day, looking at why global education is so bad and why all the global topics and issues take decades or longer to evolved – I know what does not help — paper methods, memorization, certification by “read a chapter, answer a few questions, forget”. Churning out hundreds of millions of students globally – forced to memorize formulas and stuff that is incomplete, out of date and nearly impossible to use reliably. Put the data and methods in the computer, on the Internet accessible to all humans, teach the LLMs to use the computers just like humans do, and certify the LLM wrapper programs.
Tools not words. Aim at engineers, aim at heliospheric exploration and development. Aim at feeding the world and caring for 8.1 Billion humans, and related species.
Do not force billions of humans to memorize things – give them computers who store that knowledge and can do the math and simulations – then train the humans to go beyond and train them to give clear instructions to computers. Clear instructions to computers is very very close to good communication of instructions to humans.
I am getting away from your effort here and talking about global internet issues. I like what you found, I just hate you only know how to write paper books. A few hundred million people who can use your knowledge you understand pretty well – and teach the LLM wrapper programs to use it – then most any human could say, “Computer, calculate the cutting rate with this proposed drilling mud mixture, when we dynamically modulate the drilling depth and find an optimum at today’s prices for all inputs.”
Filed as ( Mechanics and Thermodynamics are not obsolete – if you use global open methods and teach LLMs to use the computer )
Richard Collins, The Internet Foundation
