In general, AI is already capable of producing coherent text, and with hints it can even produce astrologically correct text
But we astrologers still need more, we need a step-by-step reasoning model
This is what we will probably get soon, since rumors are actively circulating about the Strawberry project from OpenAI or something similar
When we have access, we will probably be able to set up astrological algorithms for analysis and output capable of answering any question
Having immersed myself more in these LLMs, astrologers could easily get more from the existing models by becoming "power users".
Read about what it means to be a computer power user, and why power users need so much more from a computer than other people.
techterms.com
I am an Anthropic/Claude user so my "method" would be based on that model, but I am sure ChatGPT and Gemini may have similar capabilities, not to mention the myriad tools that are becoming available in the space. You could use a combination of Retrieval Augmented Generation (RAG) and Claude's Projects feature to train an LLM to become an "Expert Astrologer" -- that is, the kind that does the indepth analysis and nuanced takes using intermediate/advanced astrological principles and approaches.
Claude Pro and Team users can now organize chats into Projects. Projects bring together internal knowledge and chat activity in one place so Claude can be your go-to expert for generating ideas, making decisions, and moving work forward.
www.anthropic.com
Basically, you open up a Claude project, and let's say you want it to be an "intermediate" level traditional astrologer. You could load the project knowledge with literature from Valens, Dorotheus or "insert preferred astrologer here". Then, you could provide it with data of traditional astrological interpretations - maybe you have a catalogue of your own interpretations or you could go on a forum like skyscript.co.uk and extract the mystery chart, horary, and theoretical threads. Compile all that data and place it in the project knowledge. This would provide the model with the correct blueprint which would ground its interpretation.
You could then add a "primary prompt" to the project, using your preferred method. A fun one I encountered recently is "Train of Thought". The prompt would start of as, "Imagine three different experts are answering this question. All experts will write down 1 step of their thinking, then share it with the group. Then all experts will go on to the next step, etc.If any expert realises they're wrong at any point then they leave. The question is...". The result of all of this prep would be a model that could be your own "astrological expert group" that you could bounce ideas off of and augment you existing skill as an astrologer.
The efficacy of this approach presupposes your astrological expertise -- a relative beginner to astrology would not be able to parse sense from nonsense. This approach would make you more productive.
This could work for any type of astrology you practice -- jyotish and contemporary styles too. One way I see this really benefitting astrologers is in testing more niche techniques at scale.
