Currently, the AI world seems to be dividing into three directions: "AI" is now used synonymously with "Generative AI". In addition, "MI" is becoming established for machine intelligence including ML, i.e. machine learning. For classic AI = symbolic AI [Wooldridge (2020) The Road to Conscious Machines. P. 42] and its further development into "digital intelligence" including digital thinking, "DI" could be considered the third branch. What do you think? Is there "One AI" or how many different branches?
Translated with www.DeepL.com/Translator (free version)
The AI landscape is evolving into three main branches: Generative AI, which focuses on creating content like text and images using deep learning models; Machine Intelligence (MI), which includes machine learning (ML) and other data-driven approaches enabling machines to learn and improve; and Digital Intelligence (DI), referring to symbolic AI and rule-based systems evolving into advanced digital reasoning. While these branches highlight the field's diversity, they all contribute to the unified goal of creating intelligent systems. General AI (AGI) aims for human-like intelligence across tasks, whereas all other AI forms, such as those in the three branches, are considered Narrow AI (ANI), designed for specific tasks.
There are several different branches of one AI, because all they are describing/generalizing/extending (or trying to) the same object: the (undefined) human intelligence.
We are in a similar situation described by Marvin Minsky several (in fact, many) years ago:
https://ojs.aaai.org/aimagazine/index.php/aimagazine/article/download/894/812&ved=2ahUKEwiFkKHSmY-HAxX0tYQIHZHnA4IQFnoECB0QAQ&usg=AOvVaw0HN_9Wqj5VRysL3oPkn9C5
I would like to consider Image Recognition and Image Generation as a one more branch.
The main kinds of AI are:
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction