Playbook: LearnAI for Scientists

Your reference for using an AI assistant in research, experimentation, and scientific communication. Ready-to-run prompts for hypothesis development, data interpretation, scientific writing, and collaboration.

What This Guide Is Not

This playbook focuses on practical role-specific applications rather than habit formation. This is a research toolkit — the specific ways an AI assistant can accelerate your scientific work, and the prompts that deliver results.

Where to Practice These Prompts

Every prompt in this guide works with any AI assistant — ChatGPT, Claude, GitHub Copilot, Gemini, or whatever tool you prefer. The prompts are the skill; the tool is just where you type them. Pick the one you’re comfortable with and start today.

You don’t need a specific tool to benefit. You need the habit of reaching for AI when the work is genuinely hard — not just when it’s repetitive.

Core Principle for Scientists

Science is about generating reliable knowledge. an AI assistant can help you think more clearly, write more efficiently, and explore ideas faster — but it cannot replace scientific rigor. Every claim must trace back to evidence. Every interpretation must be yours to defend.

Use an AI assistant as a thinking partner for brainstorming, literature synthesis, and drafting. Never use it as a source of scientific fact. The moment you cite something an AI assistant said without verifying it, you’ve compromised your work.

The Seven Use Cases

1. Hypothesis Development and Refinement

When to use: When you have an observation, a hunch, or a phenomenon — but not yet a testable hypothesis.

Prompt pattern:

I'm developing a hypothesis in [field/subfield]:

Observation: [what you've noticed or found]
Existing theory: [current understanding in the field]
The puzzle: [what doesn't fit or isn't explained]
My intuition: [what you suspect might be happening]

Help me:
1. Articulate this as a testable hypothesis
2. Identify the key variables and their relationships
3. State the null hypothesis explicitly
4. List what would falsify this hypothesis
5. Suggest alternative hypotheses I should consider

Follow-up prompts:

Is this hypothesis too broad? How do I scope it for a single study?

What's the most common criticism this hypothesis would face?

What's the minimum viable experiment to test this?

Try this now: You observed an unexpected dip in enzyme activity at pH 6.5 that does not match any published kinetics model for this class of proteases. The dip is reproducible across three independent runs. Paste your data and what you have ruled out into the hypothesis development prompt. The response will suggest mechanisms worth investigating — protonation state changes, allosteric effects, buffer interactions — and the control experiments that would distinguish between them.

2. Experimental Design Review

When to use: Before committing resources to an experiment. Catching design flaws early saves months.

Prompt pattern:

Review my experimental design:

Hypothesis: [what you're testing]
Method: [describe the experimental approach]
Variables:
- Independent: [what you're manipulating]
- Dependent: [what you're measuring]
- Controls: [what you're holding constant]
Sample: [subjects, specimens, data points]
Analysis plan: [how you'll analyze results]

Critique this design:
1. What confounds could produce false positives?
2. What could produce false negatives (miss a real effect)?
3. Is my sample size/power adequate?
4. What am I not controlling that I should?
5. What's the weakest link in this design?

Follow-up prompts:

How would I design this as a pilot study first?

What would a reviewer's first methodological criticism be?

If I find a positive result, what follow-up would strengthen the claim?

3. Literature Synthesis and Gap Analysis

When to use: Understanding the state of a field, finding research gaps, positioning your contribution.

Prompt pattern:

Help me synthesize the literature on [topic]:

What I've read:
[list key papers/findings you're aware of]

What I understand:
[your current mental model of the field]

What I'm trying to do:
[your research goal]

Help me:
1. Identify the major debates or unresolved questions
2. Find where the literature converges vs. contradicts
3. Spot methodological patterns (common approaches, limitations)
4. Identify gaps where new research is needed
5. Position my work relative to existing literature

Follow-up prompts:

What's the intellectual history here? How did the field arrive at this point?

Which papers are cited by everyone? Which important work is under-cited?

What would a paradigm-shifting result in this area look like?

Critical Note: Always verify specific claims, citations, and author attributions in primary sources. LLMs hallucinate papers and misattribute ideas.

4. Data Interpretation and Sense-Making

When to use: When you have results but aren’t sure what they mean. When patterns emerge but interpretation is unclear.

Prompt pattern:

Help me interpret these results:

Experiment: [what you did]
Hypothesis: [what you expected]
Results: [what you found — be specific with numbers/patterns]
Surprising elements: [what you didn't expect]

Help me think through:
1. What do these results suggest if taken at face value?
2. What alternative explanations exist?
3. What artifacts or confounds could produce this pattern?
4. What follow-up analysis would clarify the interpretation?
5. What should I NOT claim based on this data?

Follow-up prompts:

I want to claim [X]. What evidence would I need to support that claim confidently?

A colleague argues this is just [alternative explanation]. How do I distinguish?

What's the most conservative interpretation? The most exciting?

5. Scientific Writing and Manuscript Preparation

When to use: Drafting papers, structuring arguments, improving clarity.

Prompt pattern:

Help me with this section of my manuscript:

Section: [abstract / intro / methods / results / discussion]
Target journal: [where you're submitting]
Field conventions: [any specific norms]

Current draft:
[paste your text]

Help me:
1. Improve clarity and flow
2. Strengthen the logical structure
3. Cut unnecessary words without losing content
4. Ensure the key claims are front and center
5. Match the tone expectations of the field

Follow-up prompts:

This is too long. What can I cut without losing essential information?

A reviewer says this is unclear: [specific passage]. Rewrite for clarity.

Write three alternative opening sentences for this section.

Critical Note: The final scientific claims must be yours. Use an AI assistant for clarity and structure — not for generating findings or interpretations you haven’t verified.

6. Collaboration and Communication

When to use: Preparing for lab meetings, collaborator discussions, or explaining work to non-specialists.

Prompt pattern:

Help me explain my research:

Audience: [lab group / collaborators / funding agency / general public]
Technical level: [expert / informed / lay]
Time/space: [5 min talk / 1 paragraph / poster]

My research:
[describe your work]

Key finding or contribution:
[what you want them to take away]

Create an explanation that:
1. Opens with why this matters
2. Explains the approach without jargon
3. Makes the key insight memorable
4. Anticipates common questions
5. Fits the constraints

Follow-up prompts:

Make this accessible to a smart non-scientist.

What analogy would make this concept click for someone unfamiliar with [field]?

What questions will the audience ask? Help me prepare answers.

7. Grant Writing and Funding Applications

When to use: Writing proposals, specific aims, significance sections.

Prompt pattern:

Help me strengthen this grant section:

Section: [specific aims / significance / innovation / approach]
Funder: [agency and program]
Page limit: [constraints]

Current draft:
[paste your text]

Help me:
1. Sharpen the significance — why does this matter?
2. Clarify the gap — what's unknown that this addresses?
3. Strengthen the approach — why will this work?
4. Anticipate reviewer concerns
5. Make the innovation explicit

Follow-up prompts:

The significance is weak. What framing would make this more compelling?

A study section reviewer will ask about feasibility. Help me address that.

What's the one-sentence version of this proposal?

What Great Looks Like

After consistent use, you should notice: clearer hypotheses with better-defined tests, tighter writing with fewer revision cycles, more efficient literature synthesis, and better communication across audiences. The goal isn’t for an AI assistant to do your science — it’s for an AI assistant to help you think and communicate more clearly.

Practice Plan

Month 2–3

Shift from guided exercises to independent workflows — build templates, integrate AI into your real projects, and create reusable prompt libraries for repeating tasks.

Track Your Growth

/saveinsight Hypothesis development: [description]

/saveinsight Scientific writing: [description]

Continue your practice: adapt one prompt to a current challenge, then review the result before using it.

Scientific Integrity

Using AI in research raises important questions:

• Disclose appropriately. Follow your field’s emerging norms on AI assistance.
• Verify independently. Never trust AI-generated claims without checking primary sources.
• Own your conclusions. If you can’t defend it without AI assistance, you don’t understand it.
• Maintain reproducibility. Document your methods, including AI-assisted processes.

Your reputation is built on the reliability of your work. Protect it.