Chapter Twenty — What the Map Cannot Do (2 of 2)

The research behind the table did use a machine, and the page says that too. "The guide itself was researched and written with AI assistance, under the project author's direction." It gives the author's own estimate of the labor, a year of research for one person working alone, across 2,415 web addresses on 279 websites. This book reports that estimate and does not test it. The disclosure follows the practice Chapter Fifteen described, which is to say who did the work.

Two features of the record answer Chapter Sixteen's instruments. The first is "Nothing is filled in from memory." The Media Quality Grader sets the same rule for facts from outside a document, "never from unverified impression or plausible-sounding recall." The second is that an unread answer stays empty. This book's reading is that the map keeps the Guard's lock, "A missing citation is acceptable; a fabricated one is not," by applying it to doctrine. If the documents do not say, the map does not say.

One limit the map owns should be stated as plainly as the rest. The reader edition links to the pages a profile cites. The author's full guide, which holds the complete sources and logs, "is not published, because of its size," and the page says, "to check a source, contact the author." So the check that Chapter Sixteen asks you to make, going to the source yourself, is open for every cited page. The complete record behind them is open by request only. This book's judgment is that a reader should know this before relying on the page, as the page itself asks.

Three rules follow for a disciple who uses a machine, and they are this book's own. Do not ask a general chatbot which church to join. It will give you the smooth answer of Chapter Fifteen, and the map exists because that question has a documented answer a machine cannot supply from habit. If you want a machine's help with a church's own document, give it the document and hold it to Chapter Sixteen's rule: do not grade what you have not read. And keep your save code to yourself. The page says, "The save code is your answers written as text; anyone you give it to can read them." A chat that you paste it into is someone. The map's terms also say you may not "use the work to train, fine-tune, test or populate any machine-learning or AI model or data set." This book keeps that term by describing the map and quoting short passages with credit, which the same terms allow.

Five. It is unfinished, and says so. The page is "working draft v1.21." Its appendix lists what the research leaves undecided. It does not know whether a church leaves a matter open because of the size of its tent, its trust in the Spirit, its inability to agree, or its habit of writing little. Two steps would settle more, and the page says of them, "Neither is done." It floats one idea, that freedom might reflect trust in the Spirit, and then declines to claim it. "Testing the idea would take a new field for the reason each church gives. That field has not been built." Four of the visiting questions "are the beginning of a measure of closeness, which the author has described and left unbuilt." A small church not among the 26 can apply to be added by posting its own creed, and "the first step is one pilot church."

This book keeps the map's description of itself. It is a working tool with a plain honesty about its holes, and a disciple should use it as that.

The prayer again. The page ends where it began, with the prayer. "Use it to understand other churches before you argue with them, and to pray for them by name." The prayer had a reason. Jesus asked for the oneness of his people "so that the world will believe that you sent me." The map's work is to help a believer find a place in that body. What a believer does there, and in the hours outside it, is not a matter for any table. The map ends at the door of a church. Part Six begins inside the door and walks out into the ordinary day.