Thursday, August 10, 2023


TL:DR -- This post introduces a topic of long-standing debate which pertains closely with how we ought to consider technology and choices related to it. 


Can we know it all? Can any one? Or any thing? That latter case? One might point to the new approach to AI which is doing just that, albeit the source for knowledge is the collection (of stuff, over about three decades) on the internet. A good look at that might get one to realize one huge issue: how much is believable?; or, knowable, for that matter? 

And one consequence is that we see failures and have since one major release in November (ChatGPT). Of late, the six months of experience (with multitudes making updates and supposed improvement via learning) brings out another observation: failures are increasing despite an attempt at training by the public. 

We might add, you cannot train out the crap that was originally introduced. 

Earlier, we mentioned Charles Sanders Peirce and his work which will become more usable. There are other Americans of note to bring into the discussion. We learned of one yesterday: Frederic Brenton Fitch (Wikipedia). He was a mathematical logician and is the maternal grandfather of D. Brenton Simons (Wikipedia) who is CEO and President of the NEHGS (New England Historic Genealogical Society) which we all know and love. It's been around for a while having had an illustrious start in the 1830s. The NEHGR is a phenomenal publication providing historic and genealogical information that has been verified, somewhat.  

This is brief, but Prof Fitch is known for his paradox about knowability. There are many sides to this issue. We will look at each of those. Why? They are interwoven with the debates about computing and what we have been calling "AIn't" since there is more promise than actual realization with the current approaches. 

Aside: Fitch, too, had a Paradox: (1) Possibility (blog of Bas van Fraassen). Simons mentioned him in his commentary in the summer (2023) edition of American Ancestors

One dynamic is business versus academic. Or rather, proprietary information versus science. Who owns knowledge that is applicable to us, generally?

Another aspect of S. B. Simons' background is that he has New Amsterdam heritage (van Deusen). We will look at that. Too, he has an ancestor who was born in Lawrence KS (E. L. Simons) and went to TX and became a paleontologist of note dealing with primates. Notice the drifts that we will be considering. S. B.'s ancestor visited seveal continents in his pursuit of knowledge. 

Earlier, we mentioned a Thomas Gardner descendant who was a field biologist with many publications and species named after him. 

But, technology and computing? They are inseparable. As said, we are picking this up as our operational interest with regard to all aspects some of which get little consideration in today's environment of rah-rah money chasing, silly boys rushing and breaking things, issues related to undecidability being shouted down by the bullying environment, and other things best left for later.  

Remarks: Modified: 08/09/2023

08/09/2023 --

No comments:

Post a Comment