> ## Documentation Index
> Fetch the complete documentation index at: https://nmem.guoxudong.io/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Now

> Learn to find context, check evidence, and save a result through five guided simulations.

<div className="course-meta" aria-label="Course information">
  <span><Icon icon="book-open" /> 5 lessons</span>
  <span><Icon icon="clock-3" /> \~25 min</span>
  <span><Icon icon="user-round" /> Intermediate</span>
</div>

<p className="course-intro">Learn an AI Now workflow through interactive practice in your browser. Find saved context, read a source, check what the evidence supports, and preserve the result for a fresh task.</p>

<a className="course-cta" href="/ai-now/start-a-grounded-task">Start lesson 1 <span aria-hidden="true">→</span></a>

<Note>
  All five lessons include guided interactive practice. On desktop, click **Start the simulation** in lesson 1 when you are ready; lessons 2–5 open the simulation automatically. Learn with fictional Atlas data, then apply the workflow in Nowledge Mem. The simulation uses preset replies and does not call AI or search the web.
</Note>

## What you'll build

Use the fictional Atlas case to build a traceable brief, then apply the same workflow to your own question:

<div className="learning-stages" aria-label="Workflow stages">
  <div className="learning-stage learning-stage-active">
    <span className="learning-stage-mark" aria-hidden="true" />

    <span className="learning-stage-title">Question</span>
  </div>

  <div className="learning-stage">
    <span className="learning-stage-mark" aria-hidden="true" />

    <span className="learning-stage-title">Source</span>
  </div>

  <div className="learning-stage">
    <span className="learning-stage-mark" aria-hidden="true" />

    <span className="learning-stage-title">Evidence</span>
  </div>

  <div className="learning-stage">
    <span className="learning-stage-mark" aria-hidden="true" />

    <span className="learning-stage-title">Investigate</span>
  </div>

  <div className="learning-stage">
    <span className="learning-stage-mark" aria-hidden="true" />

    <span className="learning-stage-title">Save</span>
  </div>
</div>

The fictional example asks whether Atlas should hold one cross-team planning review next quarter, using existing quarterly notes to check dependencies over the next twelve months. It does not decide how often future reviews should happen. The pilot is not approved at the start. Lesson 1 provides practice context and lesson 2 provides interview notes, so you can follow the whole case or substitute your own question.

## Start with interactive practice

Begin with [lesson 1](/ai-now/start-a-grounded-task). Continue the same Atlas task through lessons 1–4 and the drafting and saving steps in lesson 5. Open a fresh task only for the final retrieval check.

Each simulation prepares the required sample memories, sources, and earlier conversation turns. Lessons 2–5 open with progressively more context: the background check, source reading, evidence brief, and simulated research gap. These are preset examples, not your saved practice history; refreshing or replaying restores the lesson's starting state. No model setup is needed. On mobile, you can read the steps; use desktop for the interactive exercise.

## Before practising in the real app

* Complete the [Essentials course](/essentials).
* Configure and test the model AI Now uses, following [LLM Providers](https://mem.nowledge.co/docs/llm-providers). Supported local providers are also an option. Remote clients use the connected Mem server's model configuration.
* Have one relevant memory and one source, or use the fictional practice material in lessons 1–2.
* Keep the same AI Now conversation through the analysis; use a fresh conversation for the final retrieval check.

<Note>
  The outcome is a traceable recommendation or decision, including what remains unknown. You do not need to force a final answer when the evidence is incomplete. Research time may extend beyond the lesson estimates.
</Note>

## Course path

<div className="course-grid">
  <a className="course-card course-card-featured" href="/ai-now/start-a-grounded-task">
    <span className="course-card-index">01</span>
    <h3>Start a grounded task</h3>
    <p>Begin with a question your existing knowledge can help answer.</p>
    <div className="course-card-footer"><span>\~5 min</span><span className="course-card-action">Start practice →</span></div>
  </a>

  <a className="course-card course-card-featured" href="/ai-now/bring-a-source">
    <span className="course-card-index">02</span>
    <h3>Bring a source</h3>
    <p>Give AI Now the primary material behind the question.</p>
    <div className="course-card-footer"><span>\~5 min</span><span className="course-card-action">Start practice →</span></div>
  </a>

  <a className="course-card course-card-featured" href="/ai-now/ask-for-an-evidence-brief">
    <span className="course-card-index">03</span>
    <h3>Ask for an evidence brief</h3>
    <p>Separate what the source supports from what remains unknown.</p>
    <div className="course-card-footer"><span>\~5 min</span><span className="course-card-action">Start practice →</span></div>
  </a>

  <a className="course-card course-card-featured" href="/ai-now/research-what-is-missing">
    <span className="course-card-index">04</span>
    <h3>Research what is missing</h3>
    <p>Choose how to investigate one gap and check the evidence you find.</p>
    <div className="course-card-footer"><span>\~5 min</span><span className="course-card-action">Start practice →</span></div>
  </a>

  <a className="course-card course-card-featured" href="/ai-now/save-the-result">
    <span className="course-card-index">05</span>
    <h3>Save the result</h3>
    <p>Save a checked result, then find it and reread its source in a fresh task.</p>
    <div className="course-card-footer"><span>\~5 min</span><span className="course-card-action">Start practice →</span></div>
  </a>
</div>

## Choose your investigation route

Lesson 4 simulates an outcome with no public evidence so you can practise recording a gap and choosing a next step. In the real app, choose an internal follow-up for team-specific facts or Deep Research for public evidence. You can finish the core workflow with an unresolved question and a concrete next action. To practise Deep Research itself, run a research request and inspect the sources it returns, or record that the run found no relevant evidence. Planning a follow-up alone does not complete that research exercise.

## Go deeper

Read the Nowledge Mem Docs about [AI Now](https://mem.nowledge.co/docs/ai-now) and the [Library](https://mem.nowledge.co/docs/library).