How the narrative journal works.

The Siply narrative journal turns each logged drinking session into a written entry the moment your session ends. The first session you ever log generates a full AI narrative — fresh prose written specifically about your night — so you can see what the feature actually feels like. After that first taste, free-tier sessions get procedural entries built from templates; Siply Premium unlocks AI narratives for every session you log. All entries appear in a magazine-styled view inside the app.

The mechanic sounds simple. The choice to build it this way is not. Almost every drinking-tracker app on the market displays your data as charts and totals: weekly drink counts, monthly averages, BAC curves over time. These visualizations are technically efficient — they pack a lot of information into a small visual footprint. But efficiency is not the same as insight, and most users who look at a drink-count chart see exactly what they already knew.

A paragraph is different. A paragraph has to mean something. It has to choose what to emphasize, what to leave out, what counts as the story of the night. Reading is a different cognitive act than glancing at a graph, and it tends to produce different conclusions.

What does a Siply entry look like?

Every Siply entry has a date, a title, and a written body describing the shape of a single drinking session. What varies is the depth. Procedural entries summarize what you logged. AI entries synthesize it — pulling together your drinks, your intention, your mood ratings, your check-ins, and patterns across previous sessions to produce something closer to thoughtful reflection.

Here is the same Saturday night, written both ways.

Both entries are accurate. Both describe the same night. The difference is what the AI version is allowed to notice — that this was the third quiet Saturday in a pattern, that the empty-stomach observation has repeated, that your hydration habit on these nights is consistent, that the spend was within budget. None of that is invented. All of it came from data you logged. The AI's job is to pay attention.

What is the difference between AI and procedural narratives?

Procedural narratives describe one session. AI narratives describe one session in the context of all your sessions. That single distinction drives everything else.

A procedural entry sees: this drink, this time, this intention, this BAC. It summarizes that information into well-formed sentences. It is genuinely useful — much better than a chart, much better than nothing — and for many users it is enough.

An AI entry sees all of that plus your mood ratings, your hydration check-ins, your spending, your stated intention against your actual outcome, and crucially, the months of sessions before this one. It can notice that you have logged three similar Saturdays. It can mention that your hydration habit is consistent. It can observe that your peak BAC tonight was sharper than your typical pour at the same time would predict, and that this is the second time this month that pattern showed up. It is not analyzing you. It is reading back to you, with care, the things you already wrote down.

What makes the AI version richer is therefore the same thing that makes any thoughtful conversation richer: someone is actually paying attention, and they remember what you said last time. The procedural entry has a clean memory of tonight. The AI entry has a working memory of you.

There is a practical implication: the AI entries get better the more you log. A single session produces a single-session entry, regardless of tier. A month of sessions produces an AI entry that can reference the month. Six months in, the AI is writing with real context. The procedural entries do not have this property — each one stands alone, by design. AI is the part of Siply that compounds.

What data does the AI use?

The AI narrative draws on everything you have entered into Siply — drinks, timing, intentions, mood ratings, hydration check-ins, spend, and previous sessions. The more honestly you log, the more the AI has to work with. There is no reason to under-log, hide drinks, or fudge your intention rating: the entries are private to you, and the AI's value depends entirely on the accuracy of the input.

Concretely, every AI entry has access to:

• Your drinks and timing — type, volume, ABV, when each was logged
• Your stated intention — what you said you wanted from the session before it started
• Your mood ratings — how you were feeling at check-ins during the session
• Hydration check-ins — water consumed before, during, or after drinking
• Spend — what the drinks cost you, in context of your weekly and monthly budget
• Previous sessions — patterns across days, weeks, months when relevant

What the AI does not get: any personal information beyond what is in your Siply sessions. No name, no contact information, no location data, no content from other apps, no inference about who you are. Photos you upload to a session are stored with the session for your own reference, but they are not sent to the AI — the model sees structured session data only and produces a paragraph in Siply's voice. The scoping is intentional, both for privacy and for accuracy.

The tone of the writing is also intentional. Siply entries do not judge. They observe. The AI's job is not to tell you what is wrong with your drinking — it is to give you back the night you logged, with enough attention that you might notice something you would have otherwise missed. What you do with that noticing is yours.

Why a written journal instead of charts?

Charts show quantity. Narrative shows meaning. This is the underlying premise of the feature, and it is worth taking seriously.

Quantitative tracking has a well-documented limitation: people quickly become blind to the numbers they see every day. A drink-count chart that says "you had 12 drinks this week" produces a brief reaction the first time. By the third week, it is wallpaper. The numbers stop landing because they no longer carry context.

A paragraph resists this fatigue by design. Each entry is different because each night is different. The writing changes its emphasis based on what actually happened — a heavy night reads differently than a quiet one, a celebratory night reads differently than a stressed one. The narrative cannot be glanced at and dismissed; it has to be read. And reading is the cognitive act that produces the small flash of recognition we are after.

There is also a longer-term effect. Read alongside each other in the magazine view, your entries become a small piece of literature about your own behavior. You can flip through them. You can show them to a doctor, or to yourself in three months. The narrative format is harder to dismiss than the data format because it speaks the language of memory.

Where do the entries live?

Every entry lives inside the Siply app, displayed in a magazine-styled view that all users see regardless of tier. The layout — cover, contents, individual entries laid out like editorial features — is the same for free and Premium users. What differs is what fills it.

Free-tier users see their five monthly sessions, one of which (the first ever) appears as an AI narrative and the rest as procedural entries. Premium users see unlimited sessions, every one written by AI, plus the pattern insights and historical depth that come with unlimited history retention.

The magazine aesthetic is intentional, and it is universal. Drinking is one of the most heavily-photographed activities in modern life, and almost none of that photography is honest. The magazine view inside Siply is a small countermove — taking the language and craft of editorial design and pointing it at the unglamorous truth of what you actually drank, why, and how it landed.

Is the journal accurate?

The journal accurately reflects what you logged. If you log a session honestly — the right number of drinks, the right times, an honest intention, honest mood and hydration check-ins — the narrative will describe it honestly. Neither AI nor procedural entries infer or invent information. They organize what you gave them.

This means the entire feature depends on the honesty of the input. A session under-logged becomes a paragraph that under-states the night. A skipped check-in becomes an entry with less to say. Siply cannot correct for incomplete data, and it does not pretend to. There is no audience to perform for — only you, reading back the night you decided to look at.

What it can offer, given honest input, is a faithful written record of nights that would otherwise be lost to the speed of normal life. The accuracy of that record is not a feature of the algorithm. It is a feature of the relationship between you and the app, which is the real subject of the practice.

The Siply Editors write about the small chemistry and large culture of how we drink. Read more in Issue 01.