Why open source?

Be it hardware, software, experimental procedure, etc. you need to provide enough detail that someone else could reproduce your work. Only then can others build upon and advance our science. You will then benefit in turn from THEIR contributions. i.e., Nash's cooperative games.  

Is nature non-Markovian?*

 A non-Markovian process is one where future states depend upon the history or past, not just upon the current state. Jacob Barandes suggests** that "...perhaps when one takes physically fundamental non-Markovian processes and tries to shoehorn them into a Markovian paradigm, the result is quantum theory..." 

* or by emergence at higher levels of complexity?

** Pilot-Wave Theories as Hidden Markov Models, arXiv:2602.10569, 12 Feb. 2026

Robotics development

It makes sense to develop industrial robots first*, their environment is simpler, more structured. And, again, how intelligent the robot needs to be will depend upon the task and the environment.

* Before home robots, for instance. (Of course many tasks at home are automated already: dish washing, clothes washing, drying, some cooking, robot vacuums, ....)

The AI label

On the one hand "AI" has become a marketing label intended to draw in suckers. On the other hand there really are a lot of valid AI algorithms, including things as simple as linear regression. So the "AI" label can be valid.* You have to look closely.

My new hearing aids are advertised as having "AI", including advanced noise cancelation ability.** The in-office demo was impressive, we'll see how well they work over time "in the wild."

* On the other hand the label "agent" is almost always misleading. A correct definition of "A.I. agent" is something like what's given in my 1 April 2025 blog.

** My previous pair were advertised as having machine learning. I did not find that feature to be all that useful.