Intelligence is an interesting word. It's origin comes from the latin word intelligere, which means "to understand".

Every creature with a brain has some form of intelligence. One could argue even living things without brains such as plants have a sort of built in intelligence, they dont understand with their thoughts, but with their bodies in reaction to their environments. Let's go even further and entertain the possibility that planetary and solar bodies have a form of intelligence where their atoms have a built in understanding of the laws of physics.

In humans there are so many different forms and levels of intelligence, it is a pity that individuals or groups are judged so harshly by other humans on what kind of intelligence they have or dont have and how much they do or dont have of it.

I think an intelligence in an Individual that could greatly benefit humanity, could harm humanity if it became a collective intelligence. And turn that upside down, a form of intelligence that would be greatly beneficial for a collective intelligence could be quite harmful to an individual. I will not get into specifics there as there are endless scenarios and applications for those thoughts.

The word artificial also has it's original roots in latin. I have always associated that word with "fake", but on more reading, it merely means something that is created by humans, as opposed to naturally occurring in nature.

I will discuss my thoughts on what people think of when they hear the term Artificial Intelligence. And keeping in mind seeing the word intelligence as "to understand".

1.  A human created, sentient being that lives in a computer, or some human created construct. A common theme in science fiction, but still a discussion in the day to day. Would this scenario have true intelligence? Absolutely. When people talk about AI, this is the bar they often set to judge whether it is "true AI" or not. If one cannot have understanding without sentience, then this makes sense. A discussion for the philosophers...
2. A seemingly sentient algolrythim, such as the LLM models we have today. They are not sentient, but in my opinion they do "understand", inasmuch as they have been trained to.
3. An algolrythim that takes in data, and outputs data in a seemingly intelligent way, but it is obvious there is not actual understanding happening. This would be like a pre programmed NPC in a game that will react in a certain way based on the data it receives from its environment. In a way still intelligent, but not so much.
4. I don't know what goes here but it would be more of a tool, a sorting algolrythim or a search engine. Facial recognition software. You ask for specific data, or it is fed data, it returns related data. In a sense it understands, because it returns (hopefully) what one is asking for, but it is just matching data points to matching patterns in a database.

In my opinion with out highly advanced quantum computers or more advanced unknown computing system, a true version of #1 is not possible. We will probably have types of #2 that are indistinguishable from #1, but deep down even though there is intelligence, and a believable facade of sentience there is no true sentience...

If #1 were ever to exist, I don't think it would be truly artificial in the sense I don't think it would be 100% human made. Probably the computer and software built to create it would merely be a type of portal for extra dimensional intelligences to exploit.

This leads to the age old fear and doom stories about truly sentient or seemingly sentient artificial intelligence. If we are building them to be sentient by giving them every piece of data we have collected about ourselves since we have been collecting data, we surely would be at a disadvantage if they decided to work against humanity. As their understanding of us would be complete, while we would have little to no understanding about them.