Friday, November 29, 2024

Deadly consequences

TL;DR -- Metaphor? That's about all that we can do more. How to change the situation? 

--

We got used to the blue screen of death. It represented billions and billions of costs that were never recovered. And, that mode became the norm. Unfortunately. 

It's a new day. With AIn't and GenAI, we now have real death lurking (hence, the title). Oh? Is that right? You bet. We would not have to take you far to show you this and try to discuss with you how it came to be. 

Try? Yes, lots of information is cloaked. We can only surmise or try to reserve engineer which might actually be considered illegal in some situations. 

It's mess, folks. We at the Thomas Gardner Society, Inc. say, wait. Stop (like William F. Buckley argued somewhat tongue in cheek). And consider. John knows how to do this. 

So, a metaphor. This photo represent, supposedly, some bit of American life. North Dakota. Here is the photo. As said, it was unknown in provenance but appeared on Meta/Facebook. 

GenAI depiction
This is John's response on LinkedIn which points to the FB post (which might disappear which is why we captured the image). Too, there is a comment taken from the FB post. 

There's work to be done. Let's use 2025 to get it started. 

Remarks: Modified: 11/29/2024

11/29/2024 -- 


Thursday, November 28, 2024

Review, restart

TL;DR -- The past two years has seen some tremendous changes. Some of this was predicted. Much waas not. So, 2025 will be a review year as we settle into the grove after fifteen years. 

---

A recently read post provided an interchange with ChatGPT in early February of 2023. The realization of OpenAI's release was late, as that event was in November of 2022. 

Now, one month later, there was a post about recent findings of documents that gave more information about the early years. There were several posts, such as this one: Square One? Well, in our view, this is not the case; rather, there is more research to be done: Current status

Also, we started an OpEd category with which to mark post since 2010 that dealt with events which turned out to have less strong support than thought. So, this all has to be ironed out. 

This year, 2024, has been heavy into research but along several lines which has kept us busy. But, TGS, Inc. will have priority in the future since it will provide the framework for handling discussions about technology and the science behind such. 

See the Nobel posts, for an example. 

Remarks: Modified: 11/28/2024

11/28/2024 -- 

Wednesday, November 27, 2024

Nobel Week

TL;DR -- Nobel week looms. Their website will be a proper fount of knowledge about human culture through time in a technical sense. 

---

The Nobel week is coming. This year, the prizes bring attention to new types of modeling: Nobel, Medicine, 2024

Their post today, on a 1910 Prize for the 1873 doctoral thesis of Johannes Diderik van der Waals (Wikipedia), shows how far we have come in 120 years and how we can use the Prizes to track progress which will become more important as we establish a basis for computational decisioning. 

We will continue the historical look under our technology theme.

Remarks: Modified: 12/07/2024

12/07/2024 -- Changed "edit" link to the Nobel website. 

Monday, November 18, 2024

Nobel, Medicine, 2024

TL;DR -- This is the third of three gifts from the Nobel Committee. Starting with the math of machine learning and then with the chemistry work with proteins, there were heady responses with respect to the applicability of the technological insights. Ah, but wait. The Medicine prize brings us back to complexity and the reality of such. Now, one topic that will be continually on the table is this: has computing been a boon or a bane? 

--

This is the third post on the 2024 Nobel. The earlier two were: Physics; Chemistry

This post looks at the Prize for Medicine which reports on increased insights into cell regulation. 


These three can be thought of as gifts. Let's use them correctly. All things that can come from this will be discussed and utilized over the next few years. 

1. Physics - ANN and mathematical physics. I might say AIn't, but it is to the hype and misunderstanding (managers, something from nothing? - perpetual motion - need you a reminder?). Nope, the solution to the issues is right before us. It's that thing between the ears which has been allowed to go astray as it deals with complex things (yes, Mathematics - let's discuss). 

2. Chemistry - We see the hubristic side of Google (unfortunately so) tamed plus acknowledgement of lab work which is far from application (with respect to several proposed notions). At the same time, we see feedback into the thing mentioned in 1) with respect to interpretations (which is human and very much on display in the current mess) and changes that can come when empiricism works (busyness? your money focus is deleterious). 

3. Medicine - Back to reality. We now see that genetics is not under our thumb (note prizes the past few years). Insights to be gained there will play into better understanding of the computational (applied) mathematics that allows these freaky/magical moments (busyness people, look up Minsky - not Marvin, btw). Yes, complexity in another light (looking forward to seeing this generally known). 

And so, we proceed. Where we know that we are more than brain. What we need to look at further is that we can describe the necessary maturity for computation. If only, we can look at the past two decades and discuss what not to do (again) based upon our analysis.

Remarks: Modified: 11/18/2024

11/18/2024 -- 

Wednesday, November 13, 2024

Miscellany

TL;DR -- We will have several of these. We bring back a couple of images and point to several themes, one of which is technology. 

--

Just updated this image at a 2014 post (Status and focus). 

Arrival of immigrants
by year, Plymouth context

The image had moved around over the years. This was our fourth year of doing the blog for the Thomas Gardner Society, Inc. 

About this time, there were new attempts at what is called machine learning (touted as AI, known as GenAI) for several reasons. One was that the web was 20 years old or so. This caused lots of data to accumulate for both good and bad reasons. It was also 2 years after MicroSoft decided to move to the cloud pushing people off (see post (2012) thomasgardnersociety.org). In 2017, we moved the portal (to truth) to TGSoc.org to ease typing. 

Back to the image, it was used in a post (Vita Brevis, NEHGS, 2014) about what years the colonial ancestors of a group of people arrived. Another image relates to the length of the 5th generation of the Mayflower folks. When did it start and end? See the post (2019): About genearations

How long is a generation?

Finally, something about technology. The latest ACM Comm had an article on the problem of science that results from computer science not doing its job. This has to do with supporting needs of properties such as repeatability and such. The situation is dire and noway solved by GenAI which is more of a contributor to noise than not. But, we'll get into that as it's on the plate for the TGS to tackle problems of this sort. 

We have been at this blogging since 2010 and have collected a lot of data. So, with technology in mind, we need to reorganize everything. While we do that, we'll bring in changes with respect to two Thomas Gardners as suggested by the folks at WikiTree. While, at the same time, expecting that the two were one gentleman which we will discuss. 

Note our new Category: OpEd

Remarks: Modified: 11/13/2024

11/13/2024 -- 


Monday, November 4, 2024

Nobel, Chemistry, 2024

TL;DR -- We provide a link to Nobel Prize material that goes into the detail of the Chemistry prize and its computational framwork. 

-- 

Going forward,we can pay more attention to the material from the Nobel committeee with regard to the technology behind the selection. This image is from a page that looks at the history of the 2024 award for Chemistry which helps us document progress in GenAI (machine learning). 

How does AlphaFold2 work?

Advances using computational assists that reduce time and costs for experiments have been expected from the beginning, decades ago. What changes is the complexity. Fortunately, the domain of this research can have lab support as well as the use of physical experiments. 

We will discuss possible issues as we look at the topic. However, that there are those considerations on the table is beyond doubt. How we might attempt to rectify them is not so clear. 

Remarks: Modified: 11/04/2024

11/04/2024 --