Huh... You really dived into the deep research on the topic... :)
I remember my sister's final exam at the University where her topic was software that she made, which was "emulating' a doctor... It would run a series of questions about your symptoms, and in the end, give you the final "opinion" about your potential illness and give options how to treat it... That was almost 30 years ago... So, AI isn't really as new, as it looks like, but it has evolved into something much bigger...
Why this story? Well, I'm also stuck on your example of 3 professions, doctors, preachers, and lawyers... In my opinion, the first two are (at least should) act objectively, they could be easily replaced by AI... But, with lawyers, I'm not sure... Yes, they do some trivial stuff, like contracts, agreements, etc. and that could be replaced by AI... But, replacing them in the court, where they "pick" type of defense, "skip" parts that would harm their client, and do similar things, it's more tricky to implement into AI... Not impossible, but more tricky... And in the end, would that be ethical to do? Would that AI implementation bounce back and kick us in the ass?
Luckily, the AI development isn't finished and it will continue for a long time... In which direction, we will see... 😃
Naaah, I'm all around the place :D
This is very interesting perspective, this subjectivity/objectivity dynamics. And I had sort of different perspective. In each field it requires careful consideration. Let's start with medicine, some things are universaly true, like we need to move our bodies to stay healthy, we need certain nutrition, but here starts very individual eprspective. When we cross this sickness-oriented perspective it turns out each person have might have different needs to stay helathy, different diet, different exercises or drugs and habits. In that sense it needs to be very subjective oriented within some objective paradigm of keeping (or restoring) one healthy.
What you desribed is very algorithm-driven and procedural AI for health, I'd imagine system that fed with your personal data figures out what you need at the moment to improve health. There is no actual algo tree, just giant data-fed language model, black box that spits out recomendations.
When it comes to belief it seems to be really more objective with certain belief systems. But also people are on such different places of their life/spiritual path that once again crafted/personal attitude seems to makes sense for me. And I hold an opinion, that this one seems to be the furthers on the path of automation.
As for law it should operate objectively from the rule of people being equal in front of the law, however all the specific situations and details create giant tree of nuance. But objectivity comes as something that falttens individual opinions of judges. I mean, humans should stay as people creating laws, and this will be subjective process. However execution of that law should be more procedural and automatic IMO, still taking into account all the details. And what you desribed as this incourt process seems to be very replacable by current LLMs, but people involved should stay in the loop to decide the most crucial junctions.
Defnitely very complex topic, and I think that we might for a long time stay behind this tech with our everyday lifes. Which might be not really a bad thing.