How we use AI
How we use AI
Transparency matters. Here is how the analyzer works and what it is not.
Structured output, not a chatbot oracle
The engine asks for your decision, context, and constraints, then produces scenarios, dimensions, a timeline, and a score. That structure reduces vague answers compared with open-ended chat.
Your text and consent
When live AI is enabled, your question is sent to the provider only to generate the response, per their terms. We don’t run user-submitted code; don’t paste secrets into public community posts.
Ads & analytics
We may use Google AdSense and analytics with consent where required. You can read more in our Privacy Policy and Cookie/Consent banner.
Not professional advice
This tool does not replace doctors, lawyers, therapists, or financial advisers. In crisis, contact local emergency or qualified professionals.
Editorial overview
Life Decision Engine uses large language models in three places: to expand a one-paragraph brief into named scenarios, to apply a fixed set of analytical lenses to each scenario, and to draft the longer playbook articles that an editor then reviews before publication. The three surfaces have different rules, because the cost of a mistake is different in each one.
In the analyzer, the model never operates on hidden history: each call sees only the brief you typed, plus the same prompt template that anyone else gets. There is no fine-tuning on user data, no per-user memory, and no training-time use of submitted briefs. The model provider for the default tier is OpenAI's GPT-4o-mini (or the model named in the public environment), and the provider's standard zero-retention API mode is used where available.
What the AI is allowed to do
- Generate three to four named scenarios from a brief (best, base, rough, contingency).
- Apply structured lenses — money, mind, risk, opportunity — and produce short paragraphs for each.
- Produce a numerical score with a written reason that cites the lenses, not magic numbers.
- Draft long-form articles that are then reviewed, fact-checked, and edited by a human editor before publishing — every article carries the editorial transparency block at the bottom.
What the AI is not allowed to do
- Diagnose medical, psychiatric, legal, or financial conditions. The wording of every report names this explicitly.
- Recommend a specific medication, attorney, broker, or crypto position.
- Reach the open internet at runtime. Briefs are not enriched with live web search; that keeps the privacy story simple and avoids citing hallucinated URLs.
- Train on your brief. We do not opt user submissions into provider training pipelines; where the provider offers a no-train data setting, we use it.
How errors and edge cases are handled
When the model produces something that an editor flags in review (factual error, too-confident phrasing, policy-sensitive recommendation) we either re-run with a tightened prompt or rewrite the section by hand. The prompt templates live in version control and are reviewed quarterly. If you spot something that looks wrong, write to the editorial team — the contact link is at the bottom of every page — and we will correct the article and the prompt.
Frequently asked
- Will my analyzer brief be used to train an AI model?
- No. Briefs are sent to the model provider in zero-retention mode where available, and we do not maintain our own training corpus from user input. We log the high-level event (a brief was analyzed) and aggregate metrics for capacity planning; we do not log the contents.
- Why is there an AI badge on every blog post?
- Because being honest about it is the only ethical option and because Google's guidance on AI-assisted content asks publishers to disclose. See the editorial standards for the full review checklist that every article passes before publishing.
- Can I opt out of AI-touched content?
- The analyzer itself is the AI tool, so opting out means not using it. The blog posts are AI-drafted and human-edited; if you only want fully human-written content, the experts directory and the community board are the right places.
Curated by the Life Decision Engine Editorial Team · Last reviewed . Our editorial standards.