AI for Productivity & Daily Work

Writing is where most professionals spend 20–40% of their working day. AI can cut that time dramatically — without making your writing sound robotic — if you know how to use it correctly.

The right mental model: AI as your writing partner

Don't ask AI to "write an email" and paste the result. That produces generic, flat output. Instead, think of AI as a writing partner: you direct, it drafts, you refine. The best outputs come from you providing:

• The context (who you are, who you're writing to, what the situation is)
• The goal (what you want the reader to do or feel)
• The tone (formal, direct, warm, persuasive)
• Any specific points to include or avoid

Email: A proven prompting framework

Use this template as your starting point for any professional email:

Example: "Write a professional email from a senior marketing manager to our VP of Sales. The goal is to propose a joint campaign for Q4 to hit our shared revenue target. Context: We missed Q3 by 8% and both teams are under pressure. The campaign I'm proposing is a co-branded webinar series targeting mid-market SaaS companies. Key points: proposed dates, expected pipeline, ask for a 30-min call. Tone: collaborative and confident. Under 200 words."

Reports and proposals: The outline-first method

For longer documents, never ask AI to write the whole thing at once. Instead:

1. Ask AI to generate a detailed outline first. Review and adjust it.
2. Ask AI to write each section separately, providing the context for that section.
3. Review, adjust for your voice, add specific data and examples you know.
4. Ask AI to review the full draft for consistency and flow.

Making it sound like you

After AI drafts something, paste in an example of your own writing and say: "Rewrite the above draft to match this writing style: [your example]." This single step dramatically improves how human and authentic the output sounds.

Meetings are time black holes. Research consistently shows that most knowledge workers spend 15–20 hours a week in meetings — and up to 40% of that time produces no documented output. AI changes this completely.

Option 1: AI transcription tools (best for recurring meetings)

Tools like Otter.ai, Fireflies, and Fathom join your meetings, transcribe them in real-time, and produce AI-generated summaries and action items automatically.

• Otter.ai — integrates with Zoom, Teams, Google Meet. Free tier includes 300 min/month transcription.
• Fireflies.ai — stronger AI summaries, action item extraction, and search across past meetings.
• Fathom — excellent for Zoom, free for unlimited recordings, popular with sales teams.

Setup takes 10 minutes. The ROI is immediate — you stop writing manual notes forever.

Option 2: Paste-and-summarize (best for one-off or sensitive meetings)

If you can't or won't record a meeting, write rough notes during the call and paste them into ChatGPT or Claude immediately after with this prompt:

Following up after meetings

Once you have a summary, use AI to draft the follow-up email:

The compounding benefit

When every meeting produces clean, searchable notes, something powerful happens: institutional knowledge stops living only in people's heads. When someone asks "what did we decide about X in March?" you can find it in 30 seconds instead of emailing three people and getting three conflicting answers.

Research is one of the highest-leverage places to use AI. A task that used to take half a day — reading 15 articles, synthesizing findings, drawing conclusions — can now take 30–45 minutes with the right workflow.

The research workflow: 3 stages

Stage 1: Question formulation

Before you search, clarify exactly what you need to know. Use AI to help sharpen your research question:

Stage 2: Source gathering and summarization

For publicly available content (articles, reports, websites), use Perplexity AI or ChatGPT with web access to gather initial information with sources. For documents you already have, paste them directly into Claude (which handles long documents exceptionally well).

Stage 3: Synthesis and output

Once you have summaries from multiple sources, synthesize them:

Important: verification is still your job

AI can hallucinate — generating plausible-sounding facts that aren't true. Any specific statistic, quote, or factual claim that matters should be verified in the original source before you use it in a client deliverable, presentation, or report. Use AI for synthesis; use primary sources for verification.

Competitive research

One underutilized research workflow: competitive analysis. Paste a competitor's pricing page, job postings, or product description into Claude and ask: "Based on this, what can we infer about this company's strategy, target customer, and positioning? What are their likely weaknesses?"

Automation used to require coding skills. Now, AI tools let you automate repetitive tasks with plain language instructions. This lesson covers the most practical options available to non-technical professionals.

What "automation" actually means for most professionals

For most non-technical professionals, automation doesn't mean robots running your workflows. It means:

• Creating reusable prompts and templates that replace repeated manual work
• Setting up tools that run AI on a schedule or trigger (like an email)
• Using AI to handle the "in-between" steps that slow down a workflow

ChatGPT Custom GPTs

If you have ChatGPT Plus, you can create a Custom GPT — a version of ChatGPT pre-loaded with your instructions, documents, and context. Practical examples:

• Proposal writer GPT: Pre-loaded with your company's capabilities, case studies, and proposal format. Ask it to draft a proposal for any new opportunity.
• Email responder GPT: Pre-loaded with your FAQ answers, pricing, and policies. Paste an incoming inquiry; it drafts the response.
• Report formatter GPT: Pre-loaded with your brand templates and reporting format. Paste raw data; it formats the report.

Zapier + AI (no code)

Zapier connects apps and now includes AI steps. Some practical automations:

• New email in Gmail → AI summarizes it → summary saved to Notion
• New form submission → AI drafts a personalized follow-up → sent via email
• New Slack message in #support → AI categorizes it → routed to right team member

None of these require code. Zapier's interface is visual and the AI steps accept plain-language instructions.

Building your personal prompt library

The most underrated productivity tool: a document where you save your best-performing prompts. Every time you write a prompt that produces great output, save it. Within a few weeks, you'll have a personal library that lets you handle recurring tasks in seconds instead of minutes.

Individual AI tricks are less powerful than a consistent daily workflow. In this lesson, we'll build a repeatable system that compounds over time.

The 3-block daily AI workflow

Block 1: Morning (10 minutes)

• Open your AI tool. Ask it to help you prioritize your top 3 tasks for the day based on your goals.
• If you have meetings, paste the agenda and ask: "Help me prepare for this meeting. What questions should I ask? What should I know in advance?"
• Check if there's a draft email or report that needs to go out — use AI to finalize it.

Block 2: During the day (as needed)

• Any writing task: brief AI first, then edit its output.
• Any research task: use AI as your first pass, then verify what matters.
• Any complex decision: use AI to generate pros/cons, devil's advocate arguments, or alternative approaches you might have missed.

Block 3: End of day (5 minutes)

• Paste your to-do list and ask AI to help you carry anything unfinished into tomorrow's plan.
• If you recorded any meetings, skim the AI summaries and capture key action items.
• Add any new useful prompts to your prompt library.

The compounding effect

Professionals who build this kind of daily habit report saving 1–2 hours per day within 30 days. That's 5–10 hours per week — the equivalent of an extra day of productive work. Over a year, it's a compounding advantage that separates those who master AI from those who dabble.