Lesson 3.1: Choosing and Accessing AI Tools

Purpose

You've identified where AI can help in your work. Now you need to choose which tool to use. This lesson helps you evaluate options and access them, without vendor-specific tutorials (those change constantly).

By the end, you'll know the tool categories, how to evaluate which is right for you, and how to get access.

Why This Matters for Managers

Too many managers struggle with "I know AI could help, but which tool should I use?" Then they get stuck and don't take action.

This prevents:

• Analysis paralysis about which tool to pick
• Using inappropriate tools for your needs
• Wasting time learning tools that don't fit
• Security and policy violations by using unauthorized tools

The stakes: Choosing the right tool means you actually use AI. Choosing wrong means you get frustrated and abandon it.

Core Concepts

Tool Categories

AI tools fall into a few main categories. Most managers need only one or two.

Category 1: Chat-Based AI (Most Useful for Managers)

What it is: A conversational interface where you type a question or request and the AI responds. Examples: ChatGPT, Claude, Gemini, Copilot.

How it works:

• You type a prompt
• AI reads it and responds in conversation
• You can follow up, refine, iterate
• Conversation continues in one thread

What it's good for:

• Drafting (emails, documents, outlines)
• Summarizing
• Brainstorming
• Answering questions
• Explaining concepts
• Analyzing text

Strengths:

• Very flexible—can do almost anything
• Easy to learn—just type and talk
• Good at iterating (you refine, it adjusts)
• Conversational, so you can ask follow-ups

Limitations:

• No memory between conversations (each conversation starts fresh)
• Quality varies—needs verification
• Can hallucinate
• Context window limits (can't analyze huge documents)

Best for managers: Chat-based AI is probably your primary tool. It's the most versatile.

Category 2: AI Integrated Into Tools You Already Use

What it is: AI built into productivity software (email, docs, spreadsheets, project management). Examples: Microsoft Copilot in Word, Google's AI in Docs, ChatGPT in Gmail (with plugins).

How it works:

• You're in your normal tool (email, doc, etc.)
• There's an AI button or prompt
• You ask it to draft, summarize, or generate something
• It works on the content in that tool

What it's good for:

• Drafting in the tool you're already using
• Summarizing within documents
• Formatting and organizing
• Anything specific to that tool

Strengths:

• Integrated into tools you already know
• Works on content in that tool (no copy-paste)
• Consistent with your workflow
• Often included in subscriptions you already have

Limitations:

• Less flexible than standalone chat AI
• Limited to what that tool allows
• May have fewer advanced features
• Vendor-dependent (different quality across tools)

Best for managers: Use this if it's already available in your tools. Convenient. But it's often supplementary to standalone AI.

Category 3: Specialized AI Tools

What it is: AI tools built for specific tasks. Examples: AI for video editing, AI for email management, AI for scheduling.

How it works: Varies by tool, but typically specialized for one type of task.

What it's good for: The specific task it's designed for. Not general-purpose.

Strengths:

• Optimized for one task
• Often higher quality in that domain
• Focused features

Limitations:

• Only useful for that one task
• Learning another tool
• Cost (specialized tools often cost more)
• May be overkill if you rarely do that task

Best for managers: Probably not your first choice. You'll likely use general chat AI, then add a specialized tool if you repeatedly do something very specific.

Category 4: Code or Advanced Technical Tools

What it is: AI for coding, data science, or technical work. Examples: GitHub Copilot, specialized data analysis tools.

How it works: Varies, but often integrated into development environments or specialized interfaces.

Best for managers: Unless you code, skip this category. Not your primary tool.

The Tool Evaluation Framework

When choosing among tools in a category, evaluate:

Evaluation Dimension 1: Capability

What it can do: Does it handle the tasks you want? (Summarize? Draft? Brainstorm?)

How to check:

• Try the free version or trial
• Attempt your actual use cases
• See if output quality is acceptable

What matters: Does it do what you need?

Evaluation Dimension 2: Ease of Use

How hard is it to learn? Can you start using it today or does it require training?

How to check:

• Try it. If you can figure out how to use it in 10 minutes, it's easy.
• Read reviews on learning curve
• Watch a 2-minute demo

What matters: You need to actually use it. If it's too complex, you won't.

Evaluation Dimension 3: Reliability

How consistent is it? Does it work as expected, or are there lots of errors?

How to check:

• Try it on tasks you know well
• Verify outputs on a few attempts
• Check user reviews on reliability

What matters: You need to trust the output enough to use it. Unreliable tools waste time in verification.

Evaluation Dimension 4: Cost

Free tier vs. paid: Can you start free, or do you have to pay immediately?

Cost structure: Monthly, per-use, one-time?

What matters: For managers just starting, free or low-cost options are best. Pay once you know you'll use it.

Evaluation Dimension 5: Privacy and Security

Where does your data go? Is it stored? Could it be used to train future models?

How to check:

• Read the privacy policy (yes, really)
• Check organizational policy on AI tool use
• Understand whether sensitive data can be shared

What matters: Don't use free public AI for confidential information. Check your organization's policy.

Evaluation Dimension 6: Availability and Compatibility

Can you access it easily? Is it available where you work? Does your organization allow it?

How to check:

• Check organizational policy on AI tools
• Verify you can access it from your network/device
• Confirm there are no blocking restrictions

What matters: A great tool is useless if you can't access it at work.

Practical Managerial Use Cases

Decision Tree: Which Tool to Try First

Question 1: Is your organization providing an official AI tool?

• Yes: Start there. Easier to use, policy-approved, likely integrated.
• No: Proceed to Question 2.

Question 2: Do you want the most flexible general-purpose tool?

• Yes: Try ChatGPT (free tier exists), Claude (free tier exists), or Gemini (free tier).
• No: Proceed to Question 3.

Question 3: Do you want something already in your current software?

• Yes: Check if your email, docs, or project management tool has AI integrated.
• No: Go with a general-purpose chat AI.

Result:

• Most managers: Start with free general-purpose chat AI (ChatGPT, Claude, or Gemini)
• Managers with organizational tool: Start with that
• Managers wanting integration: Try Microsoft Copilot if you use Microsoft tools, or Google's AI if you use Google Workspace

Case Study 1: Choosing for a Corporation With Policy

Situation: You work for a large company with an official AI policy.

What to do:

1. Find the official policy (usually in IT documentation or HR intranet)
2. Understand what's approved
3. Use the approved tool
4. Don't use other tools without approval (data security risk)

Likely scenario: Your company has licensed an enterprise AI tool. Use that one. It's policy-approved, secure, and integrated.

Action:

• Contact IT or HR to access approved tool
• Follow organizational training if required
• Use that tool

Case Study 2: Choosing for a Small Company or Startup

Situation: You work for a smaller organization with no official AI policy.

What to do:

1. Start with a free, widely-used tool (ChatGPT, Claude, Gemini)
2. Try it for 2-4 weeks on non-sensitive work
3. Understand organization policy on confidential data (what shouldn't go into external AI?)
4. Set your own rules (e.g., "I won't paste customer data, but I'll use it for drafting")
5. Talk to leadership about whether you should formalize policy

Action:

• Pick one free tool
• Start using it on non-sensitive tasks
• Establish personal guardrails for confidential information

Case Study 3: Individual Contributor Manager at Large Company

Situation: You manage a team within a large company. Company doesn't have an official AI tool yet.

What to do:

1. Check whether your company forbids personal use
2. If allowed, start with a free public tool
3. Don't use on confidential or proprietary information
4. Advocate internally for an official tool (if you see value)

Action:

• Use free tool with caution
• Wait for organizational tool if possible
• Follow any organizational guidelines

Anti-Patterns / Misuse Risks

Misuse Risk 1: Choosing Based on Hype, Not Your Needs

"Everyone's using ChatGPT, so I'll use that even though [other tool] would be better for what I do."

Why it fails: Hype doesn't match your actual needs. You want the tool that's easiest and best for your use cases.

Misuse Risk 2: Overcomplicating the Choice

"I'll evaluate 10 different AI tools before I pick one."

Why it fails: Analysis paralysis. You end up never using any of them.

Misuse Risk 3: Using Unauthorized Tools With Sensitive Data

"The free AI tool is great for my sensitive project data."

Why it fails: Policy violation, security risk, potential data breach. Free public tools often store your input for training.

Misuse Risk 4: Expecting Too Much From Your First Tool

"If this tool doesn't solve all my problems, I'm abandoning AI."

Why it fails: No single tool is perfect. You'll learn its limits and adjust.

Human Judgment Checkpoints

As you choose a tool, ask:

1. Is this authorized? Check your organizational policy.
2. Is it easy to use? Can you learn it in one session?
3. Can I actually access it? From your network, on your device, in your role?
4. Am I starting with a free option? Unless you're certain, start free.
5. Can I protect confidential data? Understand what can and can't be shared.

Responsible AI Considerations

Following Organizational Policy

Many organizations have AI policies. Follow them. If you're uncertain whether a tool is allowed, ask IT or HR.

Protecting Confidential Information

Free public AI tools may store or use your data. Don't paste confidential, proprietary, or sensitive employee information into unauthorized tools. If uncertain, use only approved tools.

Starting With Free Tiers

Free tiers let you experiment without commitment. Use them to learn whether AI is actually useful for you before paying.

Being Vendor-Agnostic

Don't fall into proprietary lock-in. Learn the concepts (prompting, verification, application), not just one tool. Concepts transfer between tools. Tools change.

Practice / Reflection Prompts

1. Policy Check: What is your organization's official policy on AI tools? Where did you find it?

1. Needs Assessment: For the use cases you identified in Chapter 2, what capabilities do you need in a tool?

1. Trial Selection: Pick one tool you'll try for two weeks. What use cases will you focus on?

1. Ease of Use: Can you learn your chosen tool in 15 minutes? If not, is it worth the learning time?

1. Data Safety Plan: What's your rule for confidential information? What can you paste into the tool? What can't you?

1. Backup Plan: If your first tool doesn't work out, what's your second choice?

Key Takeaways

1. Start with a general-purpose chat-based AI. Most versatile for managers.
2. Check organizational policy first. Don't violate security guidelines.
3. Start free and simple. Use free tiers before paying.
4. Evaluate based on your needs, not hype. What will you actually use?
5. Protect confidential data. Know what information can/can't be shared.
6. Give it 2-4 weeks before deciding. Learning curve is real.