Research and Comparison Tasks — AI for Work & Productivity

You'll learn

The organizer-not-source rule for AI research
Criteria-first thinking: decide what matters before you compare
How to build comparison tables that actually help you choose
Which facts to verify — and where
Your personal research-and-decide routine

01Section

AI organizes. You verify.

Research is where AI is most tempting to misuse. Ask "which project tool is best?" and you'll get a confident, specific, beautifully organized answer — with prices that changed last year, features that were discontinued, and maybe an option that doesn't exist. All delivered in the same polished tone as the true parts.

So this whole lesson runs on one rule: use AI to structure the research — gather, organize, compare — and treat every fact it supplies as unverified until you've checked it somewhere real.

The rule

AI is your research assistant, never your source. It organizes; you verify.

AI Safety & Privacy path

Quick knowledge check

AI hands you a detailed comparison of three tools, with prices and features. How should you treat it?

AAs accurate — the specific details prove it did researchBAs a useful structure whose facts need checking against original sourcesCAs completely uselessDAs accurate for the most popular option only

Checkpoint

02Section

Decide what matters before you compare

Weak research starts with "which is best?" — because "best" quietly hands the decision criteria to whoever answers: the AI, a marketer, a stranger's review. Strong research starts with you naming your criteria: budget, must-haves, dealbreakers, who has to live with the choice.

Criteria outsourced

"Which CRM is the best?" You'll get the most-mentioned option — optimized for the average of everyone, which is to say, for no one.

Criteria owned

"I'm choosing a CRM for a 4-person team. My criteria: under $50/month, works with the email we already use, learnable in a day, exports our data if we ever leave. Compare options against these." Now every row of the answer serves your decision.

Key takeaway

Criteria are the decision. Choose them yourself, before AI enters the room.

Checkpoint

03Section

Ask for a table

The comparison table is the workhorse of this whole skill: options as columns, your criteria as rows, one glance to see where the tradeoffs live. Two instructions make it trustworthy — tell the AI to write "verify" in any cell it isn't sure of, and ask for the list of facts to check before deciding.

Even better: paste your own gathered material — pricing pages, notes from demos, snippets from reviews — and have the AI organize that. Then the table is built from things you've already seen with your own eyes.

Comparison table

I'm comparing options for a decision. My options: [LIST OPTIONS]. My criteria, in order of importance: [LIST CRITERIA]. Build a comparison table with my criteria as rows and the options as columns. Where you're not certain of a fact, write "verify" instead of guessing. After the table, list the specific facts I should verify before deciding, and where I'd check each one (official pricing page, documentation, recent reviews). Do not recommend a winner — lay out the tradeoffs and leave the decision to me.

Quick knowledge check

Why tell the AI to write "verify" in cells it isn't sure about?

AIt makes the table easier to readBOtherwise it fills the gaps with plausible guesses that look identical to factsCIt makes the response arrive fasterDTables require a placeholder in every cell

Checkpoint

04Section

Verify the facts. Make the call yourself.

You don't verify everything — you verify what would change the decision: the price, the dealbreaker feature, the compatibility claim. Check those in original sources: the official pricing page, the documentation, recent reviews from real users, or a direct question to the vendor. A comparison built on invented prices produces a confident decision about a world that doesn't exist.

And then the last step stays yours. Even with every fact verified, don't ask AI to make the call. It doesn't carry the consequences, doesn't know your team's patience for new tools, and didn't sit through the demo that gave you a bad feeling. Have it argue against your leading option — that's a genuinely useful stress-test — but the decision is yours.

Plain-language definition

The decide-yourself rule — AI can gather, organize, compare, and argue both sides — but the person who lives with the outcome makes the choice. Delegating the decision means letting a tool with no stake pick for you.

Privacy note: Comparison work often touches internal numbers — budgets, vendor quotes, contract terms. Round them or leave them out: "under $5k" compares just as well as the exact figure from the quote.

Pause and think: What decision are you circling right now — a tool, a vendor, a plan? What are your top three criteria, honestly?

Checkpoint

05Section

Your research-and-decide routine

This is the path capstone, so let's assemble the whole thing into a routine you keep — six steps that work for choosing software, picking a vendor, or any digest-and-decide task:

Question: Write the decision as one sentence: "Choose X for Y by [date]."
Criteria: List what matters, in order, dealbreakers first — before any AI is involved.
Gather: Collect real material where you can: pricing pages, docs, notes, reviews.
AI-organize: Hand over your criteria and material; get the table, the "verify" gaps, and the checklist.
Verify: Check the decision-changing facts in original sources.
Decide: You make the call — after asking AI to argue against your favorite, if you want the stress-test.

Why this is the capstone

This routine is the whole path in miniature: AI supplies structure and speed; you supply judgment and keep the decision. That division of labor is what you've been practicing for six lessons.

Checkpoint

Prompt exercise

Run your research-and-decide routine on a real decision

Pick a real decision you're facing — a tool, a purchase, a vendor, a plan. Copy this prompt into ChatGPT, Claude, Gemini, Copilot, or whichever AI tool you have access to; the website doesn't run AI itself. Then save the prompt somewhere permanent — it's your reusable routine, and it's the artifact this path leaves you with.

You're my research assistant for a decision. Walk me through my routine one step at a time, waiting for my answer at each step:
1. QUESTION — Ask me to state the decision in one sentence.
2. CRITERIA — Ask me what matters, in order, and what my dealbreakers are. Don't suggest criteria until I've given mine.
3. GATHER — Ask me to paste any real material I have (pricing pages, notes, reviews). If I have none, say so and continue.
4. ORGANIZE — Build a comparison table from my criteria and material. Write "verify" in any cell you're not certain of.
5. VERIFY — List the facts that would change my decision and where to check each one. Wait while I check, then update the table with what I confirm.
6. DECIDE — Don't pick a winner. Summarize the tradeoffs, briefly argue against my leading option, and remind me the call is mine.

Reflection: You just went from question to decision without once trusting AI as a source. Save the prompt — it's yours now. Six lessons ago you were summarizing an email thread; this is the same principle grown up: AI does the structure, you do the judgment.

Quick check

4 quick questions — no pressure

There's no pass or fail here. Answer them all, and we'll show you the answers either way.