Where Do We Even Start with AI?

A Field Guide for Businesses on the Edge of Adoption

If you've been asking yourself, “How do we actually start using AI in our business?” — you're not alone. Every week, I meet smart, capable business leaders, managers, and technical leads who feel the same thing: they know AI is important, they know it's not a passing trend, and yet — they don't know where to begin.

The real problem? Most of the education out there either goes too high-level (“AI will change the world!”) or dives deep into tech jargon that assumes you're already building custom LLMs in your garage.

It's like trying to teach someone to cook by handing them either a Michelin star philosophy book or a molecular gastronomy chemistry set—when all they wanted was to make dinner.

This post is for everyone in the middle. The practical builders. The forward-looking decision makers. The people who know something's gotta happen, but aren't quite sure what the first move should be.

1. Why “AI” Isn't One Thing

Let's clear something up: AI isn't a product. It's a toolbox. Inside that toolbox are a lot of different instruments:

  • LLMs (Large Language Models): Think ChatGPT, Claude, Gemini, Mistral. These are general-purpose tools that can summarize, generate, and explain language.
  • Agents: Task-focused systems that reason through multi-step goals (e.g., “book me a trip” or “analyze this spreadsheet”).
  • Retrieval-Augmented Generation (RAG): AI working with your own data or documents in real-time.
  • Automation tools: AI-driven emails, customer support, help desk routing, report generation.

Understanding that AI has components is key to knowing what you can do with it.

2. Start Small: Entry Points That Work

Here's the good news: you don't need a massive budget or a PhD to get started. Many businesses start with one of these:

  • Process automation: Use AI to summarize meeting notes, auto-tag tickets, or generate routine reports.
  • Customer experience: Build an AI assistant that answers common support questions.
  • Internal tools: Connect AI to your knowledge base so your team can get quick answers.
  • Sales and marketing: Draft emails, rewrite proposals, summarize leads or meetings.

Pick a real problem your team deals with every day. Start there.

3. The Three Mindsets for Adopting AI

Not every AI adoption strategy is the same. You need to know your intent.

The Efficiency Mindset:

  • Save time and reduce cost with AI.
  • Back office workflows.
  • Customer service optimization.
  • Email management.

The Exploration Mindset:

  • Discover new uses for AI that haven't been tried before.
  • R&D and prototyping.
  • Marketing and creative generation.
  • Strategic brainstorming.

The Empowerment Mindset:

  • Upskill teams.
  • Create internal copilots.
  • Personalized decision support.

4. Build or Buy?

Should you buy an AI tool or build your own? Both paths are challenging:

  • Many AI tools are too generic, not customizable, or focused on tech demos rather than business outcomes.
  • Building your own is expensive, requires skills, and often slower than expected.

Look for semi-custom options — platforms with APIs, plugin support, or low-code builders — so you can tune tools instead of just using them.

5. Your First AI Project Might Not Be Sexy — But It Will Matter

The most impactful AI projects often start with unglamorous internal problems:

  • Cleaning data.
  • Connecting isolated systems.
  • Logging decisions and outcomes.
  • Reducing time-to-answer for your team.

These clear wins build trust and create case studies to scale AI use.

6. Conversation Starters (for Your Team or Boardroom)

Try these prompts to spark leadership discussions:

  • What internal process could we automate if we had 1,000 interns?
  • Where are we wasting time hunting for information?
  • What decisions could improve with better context or summaries?
  • Where do we spend time on documentation, formatting, or rework?
  • How can customers help themselves using our own content?

Final Thoughts

The future of AI adoption won't be decided by hype or headlines. It will be shaped by those who roll up their sleeves, tackle real problems, and use AI as another tool in their toolkit.

Start small, plan well, and grow with purpose. Need help finding your first step? I'm here.

Remember: How do you eat a frog? One bite at a time.