Simmons Foundation, on training AI

TL;DR -- Jim Simmons has been almost a myth due to his success hidden behind the walls of the "hedge" phenomenon. However, his Simmons Foundation is non-profit and open. And, their approach is amenable to what we want to focus on, define, and accomplish. An example of the Foundation's work is shown. 

--

I have followed Simmons' work from a distance for a while. Jim Simmons is an example of the mathematics and physics interest in things economic and financial which needs some scrutiny. We have written on that approach which leads to a non-sustainable economy several times. However, Simmons founded a non-profit: Simmons Foundation (SF). 

That caught our attention, as an example of approaches that merits attention and as we noted with the Gairdner Foundation's work. The Gairdner focus is older and bio-medical. SF has a broader perspective which aligns with our technology thrust. 

A news article got us to look at SF a little closer today. This work caught our eye, in particular. But, their overall thrust is appealing. This is an example. 

• 
  New Datasets Will Train AI Models To Think Like Scientists - example of curating the data prior to training. As well, the specifics will be interesting as this approach would allow heavily accurate machine learning (ML) systems that would be locally expert. 
  

Okay, given that, let's get back to the foundational issues. We use AIn't for very good reason. Too, the emphasis ought to be ML. And, "think like scientists"? That is hugely debatable, and without resolution which we intend to show. 

What is missing is "adult" leadership in the equations underlying decisions plus any understanding that by necessity is imperative. We will get to the expected resolutions, eventually. How and why will not be subject to the extreme attempts at control as we see coming from the "numeric" camp. 

Most of the material has been touched upon in our posts abetted by links ot other sources. So, the work will be more than indexing and determining proper generalizations. We have to resolve open issues which are not new, generally. The means for proper research is there after decades of maturing systems brought by the efforts of many. Processes need to follow suit and will through time and effort.  

Remarks: Modified: 01/13/2025

01/13/2025 --