World Models

A.s.a. H. learns models of the world, itself, and itself in the world. As one simple early example see my blog of 4 September 2024. 

AI Pluralism again

Rashid Mushkani has a new paper on AI Pluralism.* One early example of A.s.a. H.'s value pluralism is presented in my blog of 19 February 2011. Examples of A.s.a. H.'s ontological pluralism are work like that in my blogs of 22 April 2013 and 12 September 2013. AI specialists also may have various differing sensors, outputs and actuators, architectures, etc. and perform/fulfill quite different tasks/functions/roles.

* Al Pluralism and the Worlds It Misses, arXiv: 2606.16167v1, 15 June 2026, see also my blogs of 6 April 2026, 1 January 2026, 28 June 2024, 1 June 2024, etc.

Flint Hills Digital Campus

 I have very mixed emotions regarding the gigawatt 1000 acre hyper scale data center being proposed for Emporia. On the one hand I always want larger computers. On the other hand the power and water demands are so high that I can't help but think that they are doing it wrong.

Data centers in space are not a viable alternative at the moment. They would have to come down in price by an order of magnitude in order to be competitive.

A.s.a. H. hardware upgrade

Although A.s.a. H. robots may make use of cameras* I have done very little original work on computer vision, there are so many other researchers doing that. But I have now upgraded my capability a bit with the addition of a Nvidia Jetson Orin Nano.

* for example see the concept structure in my 4 September 2024 blog  

Specialists

My A.s.a. H. A.I.s are specialists* and so am I. So there are some things that my A.I.s are more expert at than I am** and there are some things that I am better at than some of my A.I.s.***

* As one means of coping with complexity. 

** Pocket calculators have long been better and faster at arithmetic than I am.

*** I am more physically coordinated than some of my A.s.a. H. mobile robots. (but not some surgical robot)

Omnipresent bandwagon phenomenon

 Leaving aside the financially inspired hype I have observed a strong bandwagon phenomenon that was present during my work in plasma/fusion, higher education, and computer science/artificial intelligence. Popularity outweighed individual critical judgement. It strongly influenced what was publishable and what got funded.

See "bandwagon effect" on wikipedia.

Vibe coding

I agree with Linus Torvalds, "I think it's fairly positive that vibe coding means that anybody can get a computer to do something that they otherwise couldn't have done. But from a maintenance standpoint, vibe coding is horrible, horrible. Because you don't understand what you're doing."

For a developer the remaining work of formal specification, testing, and debugging is frequently more difficult than the actual coding, and a lot less fun. 

I do use LLMs as a natural language database to retrieve example code. This is as a supplement to my code library.

See also my blog of 20 March 2024.