After the reissue of Waite's deck and the boom of the 1970s, Tarot split into currents: traditionalists reconstruct historical decks, the psychological school (James Wanless, Rachel Pollack and her '78 Degrees of Wisdom,' 1980) reads cards as a mirror of the psyche, and artists create authorial worlds — from minimalism to photocollage.

A contemporary deck is always an answer to 'for whom and for what.' There are decks for therapeutic work, for queer communities, for shadow work, for meditation, for pure aesthetics. When choosing a deck, ask not 'is it correct' but 'what task did the author set, and does it match mine?'

The guidebook is part of the system, not an advertising leaflet. A good guidebook explains the choice of every image; a weak one lists generic phrases. If the author cannot explain why this symbol is on the card, the reader is left alone with a pretty postcard.

Apps have changed practice: the card of the day arrives in the morning, the journal keeps itself, the spread is saved and searchable. Digital shuffling is more honest than a tired person's hand — the generator does not care which answer you want. Lifloris stands exactly at this point: an instant local card of the day, AI interpretation with the context of your question, and a library of meanings at hand.

The boundary runs along responsibility. AI can assemble a coherent text from cards and a question, but it bears no responsibility for promises and cannot feel when the person at the table is unwell. Use digital readings as practice and a mirror; in acute states, a living person and professional help matter more than any algorithm.

A useful skill of the AI era is dissecting machine text. Take a generated interpretation and mark it with colours: observation from the card, traditional meaning, assumption, advice, generic phrase. After ten such dissections you will see the structure of any reading — machine or human.

Contemporary readers work in person, over video, in apps and social networks. Publicity raises the requirements for confidentiality, ad labelling, image copyrights, and clear service boundaries. Beautiful content does not cancel responsibility for promises.

Learning is useful in an environment where you can compare readings and get concrete feedback: which card detail supports the conclusion, how the position changes the meaning, where an unverifiable claim sounds. Reputation is built not on mystical infallibility but on clarity, ethics, and the ability to acknowledge the method's limits.

Five criteria: structure (are the 78 cards and suit logic preserved), readability (can scenes be read without a cheat sheet), cultural context (do you understand the world the images come from), the guidebook (does the author explain their decisions), purpose (does the deck's intended use match your practice).

Caution is needed with decks that borrow sacred images of living traditions without their carriers' participation. The issue is not prohibition but precision: an image torn from context stops reading — you see decoration instead of symbol.