Josiah Willard Gibbs

TL;DR -- Josiah is a descendant and a well-known contributor to the development of a major theory in physics: thermodynamics. We will look at both his work and his pedigree. 

--

In our studies related to the Science of Technology, we will be running into many names. Some will not be American (such as Noether). Some will be American, and, of those, we will look for New England associations. The series on the Harvard Presidents is an example. We started this when we ran into a New Yorker article on James Bryant Conant. That got us to looking at all of the Presidents. 

Recall, one motivator for this work related to technology is machine learning (ML) masquerading as artificial intelligence (AI). But, in general, we want to follow the evolution of STEM along with the advance of the U.S. whose 250th is coming up. 

The Nobel prizes this year were a motivator,as well, with the focus on computation: Physics (neural nets); Chemistry (protein folding); Medicine (microRNA).  

Now, back to New Englanders, some will be descendants of Thomas Gardner and Margaret Fryer. We ran into one this week. Josiah Willard Gibbs (Wikipedia; WikiTree). How did we get to him? In our technology studies, we are looking at the foundations of science. In that regard, thermodynamics is important. There are many other types of dynamics. Right now, we are considering those related to science and engineering but will broaden the scope as we go along. 

A well-known treatise on the subject was written by Max Planck (archive.org). Planck references several prior researchers, such as van der Waals. But, Gibbs is a frequent reference. On looking further, we found his association with James Clerk Maxwell (well-known contributor to mathemattican physics) and one of his detailed books on entropy that published his experimental work: On the Equilibrium of Heterogeneous Substances. We will use this work to leverage several discussions about knowledge and machine learning. 

So, Josiah will feature in future posts. He is a descendant of Thomas and Margaret via their son George. He is mentioned in later edition of Dr. Frank's book (1933: pg 63). 

Postnote: It is fortunate for us that we can use a Thomas descendant to scrutinize machine learning along several frameworks in our technology research. As such, this focus will cover both the U.S. and European contributions, indexed according to the times of the Nobel prizes with Planck's efforts as central to the discussion. 

Remarks: Modified: 12/14/2024

12/14/2024 --  

Science of technology

TL;DR --  We will be more technical and explicit about such in 2025. Our slogan: Unsafe at any locale or use. What? Yes, AI. 

---

Note: this is cursory and will be edited over the next few days.

As 2024 winds down, we are closing out some activity while preparing to continue our work in 2025 and beyond. We have several irons in the fire and stop to look at two in this post. 

• The 250th is closer but has been approaching with notable reflections for a while. Lexington/Concord will come into view in April of 2025. At some point, Bunker Hill will be the focus. We have been researching an area of Los Angeles, CA that was named Bunker Hill West. 
• Gibbs of New England was a serious thermodynamics researcher. He worked with Maxwell. Our emphasis, in part, is technology. So, let's look at science's role. 

.Gibbs and Maxwell  

Maxwell's 3D model (scuplted by hand) of Gibbs' thermodynamics equations in action. ... 

As many have said, AI is not safe. We can think back to the time of Nader and his look at the Corvair. That was a case of design trumping engineering. We have seen that over the decades of progress. The past two decades of computing (as driven by the west coast) can be seen in this light. 

Gibbs and Maxwell were way before computing came on the seen so their work is of interest to an approsch that looks at human abilities. This is to lay out some basis to establish a means of comparison. 

Oh yes, statistical mayhem? That is part of the problem. 

Also pedagogy will get a look or two. Education seems to forget that science is provisional. Gibbs is a perfect example as he laboriously wrote 600+ pages of calculations which were related to numeric review of ideas that were about measurable processes and results. As opposed to? Russell and Whitehead's long and drawn out look at logic which they stopped. 

Basically, if we looked at the guts ot AI, we would see millions and upon millions of lines of code. Perhaps, we ought to use billions (there isn't an exasy way to establish this - but, it will be important in the future with the obvious waste of the AI approach of untethered machine learning. 

Back to human learning which is culture. Look at the photo from a 1946 cover of a Fortune magazine issue: Maxwell and Gibbs.  

Remarks: Modified: 12/14/2024

12/14/2024 -- 

Nobel, 1901

TL;DR -- Let's spend some time using the Nobel prize to track the history of advances in science. After all, people do the science. And, some factors related to people come out of genealogical and historical (cultural) studies. 

--

We have paid more attention to the Nobel prizes this year than ever before. This might become a habit since the 2024 prizes referenced, indirectly, the bugaboo of the modern world, AIn't. The Physics Prize was for machine learning; the Chemistry Prize was for protein modeling. In looking at the Nobel material, it seemed obvious that the material presented about the work of het honoree was remarkable. We can follow the prize over the year to obtain material for analysis.  

As an aside, we mentioned a Balch descendant who was a Peace Prize honoree: Emily Greene Balch. But, we need to get more particular. 

Looking at Physics, first, the inception of the prize was in 1901: Wilhelm Conrad Röntgen. Here is the link to the Award Ceremony Speech. There are many years to cover. 

With the 2024 Prizes for Physics and Chemistry dealing with computational approaches which is our technology interest, we have a natural categorical framework with which to start. We will not do a linear search but will follow some interesting parallels. For instance, the 1918 Prize for Physics went to Planck for his work in quantum mechanics. Who we have referenced him in several place. In 1922, Einstein got the award for his work with blackbody radiation in 1905. Einstein presented his relativity theory in the meantime; the committee felt that there was not enough proof for an award. We'll go more into that later. 

This thematic bit of research will cover a lot of ground. Well, with respect to machine learning, we are talking the sum total of knowledge that can be handled using computational means. 

Switching to the Chemistry Prize, we thought to look at how many awards dealt with the computational in Chemistry. We expect this in Physics. Wiley published an essay that lists prizes, both Physics and Chemistry, that will be of interest to us: The Nobel history of computational chemistry. A personal perspective. 

We looked an early Chemistry prize, 1910, which was for an 1893 thesis: Nobel Week. The theme was remarkable for several reasons, one of which is the history of science as it is unfolded through human effort.Johannes Diderik van der Waals (Wikipedia) showed that elements are molecularly the same as they go through phases. We will get more into that topic as we go along. 

A final note provides a summary by Gemini of Google via Chrome. After all, the Nobel committee opened this door; we ought (actually, must) to use it. 

Courtesy, Gemini of
Google via
Chrome

Remarks: Modified: 12/07/2024

12/07/2024 --