From A Reader, Replying To My Recent Comments On China And, Unrelated, AI -- July 6, 2025

I have contacted you several times in the past on sundry matters. I am the  former ND guy who has lived in .................. I retired ....... five years ago from a technical position (........................). I had a couple of comments that I wanted to share with you.

I have been following the affairs in China for quite some time.  ...............  In the last year, Xi's status has diminished greatly and many of his appointees to positions  and supporters have been purged.  He has had significant health issues but the real status of his health have only been guessed at.  The supposition is that is to be a change in government at the end of summer into fall.  There are many blogs and podcasts on China, but the one I trust to not just follow rumors is:

Lei's Real Talk            https://www.youtube.com/c/LeisRealTalk

She does several videos per week touching on CCP matters and current status of China.   You may also be interested in past videos pertaining to China's population is no where near the stated number.  There is much evidence that their population is more in line with 700 to 900 million rather than state reported figure.  There are whole towns that have been razed since there were very few people living there.  The economic slowdown coupled with tariff threats has wreaked havoc with their manufactureing output. Many people are being paid a fraction of their normal wage since the companies they work for are in a cash crunch.

In another matter,  I had worked in AI since the middle 80's. One of my roles .....was to generate software for technical purposes such as analysis of technical data since there was no commercial software available at that time.  While there is great potential for AI to be used, much of the current applications are a lot of hype. The current versions of AI to generate applications is at best glorified pattern recognition. That is why it has been used for photo / image identification in security applications.  There is no intelligence in AI, it is just very good at recognizing patterns that are similar or have connections.  The current attempts to use information on the web to generate results is a problem in there is no check for any validity. The whole analysis is circular and problematic.  False information is given just as much weight as true information and the results are only based on frequency.  If 51% of the information used is false or not quite true, the result provided will be coming from the false information.  .......  I was working with several universities to generate new approaches to analyzing these types of data. When used properly, the techniques are quite powerful.