Thursday, June 29, 2023

Getting technical

TL;DR -- Technology covers a whole lot of territory. STEM is an universal theme. We looked at Bard of Google with respect to its abilities, in particular looking at the importance of knowing about computational lenses, truth engineering, and psychether. ChatGPT of OpenAI had little to say. 


We have had several posts on technology which covers a lot of area. It is the main focus for the future along with  human relations. Our interest now is the full scope. So, we are interested in science and experiments, including field work. We have looked at how rivers could be a boon and a bane during travel or from other aspects. Transportation is always of concern. The latest aspects of technology deals with STEM related work, especially those related to advanced computing

We will avoid the use of AI (and AIn't) and talk algorithms, instead, in their myriad of ways of being. Today, we spent time with Bard of Google. Earlier, we mentioned ChatGPT in several posts. While we were teaching ChatGPT about Thomas Gardner, a researcher was digging into the records at Sherborne Dorset. Those two bits of work will be integrated after a little more research. 

So, today's activity culminated work with Bard that had to do with three topics: computational lenses, truth engineering, and psychether. After some training, the final summary by Bard is found in this post: Bard rocks. Related posts can be seen on the Blog Archive for June 2023. 

If one overlooks the influence of current themes, there is a cohesion here. Bard also provided an example of how the three might be used in one context. Here is an attempt (by Barfd) at defining emerging concepts. 
  • Computational lenses are tools that allow us to view data through different perspectives. They can be used to identify patterns and trends that would not be visible otherwise. For example, a computational lens could be used to track the spread of misinformation on social media or to identify areas of bias in a dataset.
  • Truth engineering is the practice of designing and deploying technologies that can help us to identify and verify truth. This could involve developing new algorithms for detecting fake news or creating platforms for fact-checking.
  • Psychether is a field that studies the psychological effects of technology. It can be used to understand how our emotions, thoughts, and behaviors are influenced by the devices we use and the data we consume.
One of the issues related to the attempts at Bard (and ChatGPT and a number of other examples) revolves around a drive for "omniscience" which forces the mode of training against everything in order to be able to answer questions about everything. And, that will not work. We see it now with the problems related to the release of these models. The Large Language Model may allow nice interchanges with people, but it's based upon summaries and cannot handle details. 

As the saying goes, the devil is in the details. 

Did this choice have to be sustained? Well, it cannot be. But, switching to more locally derived data sets would allow some serious work to be done. We have already shown that. There will be more later. 

  • computational lenses deal with supporting major human endeavors with advanced systems as we have watched evolve over the past five decades. It is a metaphor with a particular focus in mind, namely human-machine interchanges. Other approaches that we have seen deal with things like computational physics where we have a twin relationship (twixt the object of the phenomenal world and its virtual/digital twin). Physics modelers and simulators abound as do many others, as we see in medicine. 
  • truth engineering was defined in a white paper at Boeing by John and was motivated by work in knowledge based engineering where very difficult design decisions were made with the support of computational resources. Engineering likes to use tradeoffs for the hard problem of balancing factors. The means for doing the work vary, but the viability of the process requires professional engineers in the loop who provide expertise and value judgment. Issues related to this concept are at the core of the discussion and experimentation with regard to machine-oriented decisions. 
  • psychether is a portmanteau created by John to suggest the importance of the ideas of Jung (MD, psychiatrist) and Einstein (Relativity), as they both were instrumental in their area of expertise with regard to grappling with universal issues, for resolving current dilemmas and to bring more attention as to how this may be the case.   

As noted, Bard was great. On the other hand, ChatGPT of OpenAI had little to say about any of these topics let alone attempt an integration. The purpose here was to use the terms in a conversational mode as allowed by the LLM in order to build coherent phrases as needed by the various viewpoints to which the concepts would apply. 

Again, our focus will be on technology with an undetermined, as yet, focus. My leaning is to advanced computing and its ubiquitous set of issues. 

Remarks: Modified: 08/08/2023

08/06/2023 -- We're getting GB XIII, 1 together. So, here is a summary of some tech work. Some of the material came from a session with ChatGPT. I marked these with italics, today, and added an image from the first search on Thomas Gardner. Notice, in the "Disclosure" there is my brief overview. This list is in reverse order by time. "psychether" was coined in the mid-1990s in a discussion. "truth engineering" was about 2000 when there was post Y2K efforts at defining the future. Y2K? Remember the fear of the 2000 rollover of the computer calendar. Then, "computational lenses" was about 2003 in reference to my work dealing with computational mathematics where we worried about the maintaining of accuracy and veracity through transforms needed for various algorithmic processes. The world is not homogeneous. The situation for the last was a presentation at a technical conference. 

No comments:

Post a Comment