Artificial Intelligence (AI) has been around for years, but if you’re not technical, it can feel overwhelming. How does AI actually work? And how can you use it to scale your business?
AI in Plain English: What It Does and Why It Matters
AI is about teaching computers to handle tasks that typically require human intelligence, like spotting patterns, making decisions and generating text. For a deeper dive, check out Ben Thompson’s analysis on Stratechery and Tomasz Tunguz’s insights on AI adoption.
Here’s how businesses are using AI right now:
- Automating tedious tasks – Scheduling meetings, generating reports, and responding to customer inquiries without human intervention.
- Turning data into insights – AI can process massive datasets and extract key takeaways to help you make smarter decisions. Upload an Excel file and ask it to conduct analysis you’d normally do yourself.
- Generating high-quality content – with the right prompting, AI can save you time writing high-quality marketing copy to investor emails. According to a Boston Consulting Group study, businesses that successfully scale AI see significant revenue increases, with modest investments generating up to 6% more revenue and larger investments tripling the impact to 20% or more.
What Are Large Language Models (LLMs)?
Large Language Models (LLMs) are essentially AI that’s really skilled at understanding and generating text.
If you’ve ever chatted with ChatGPT, Gemini (formerly Bard), or Claude, then you’ve used LLMs.
Where Do LLMs Get Their Data?
A common myth is that LLMs “scrape the entire internet.” In reality, they’re trained on a mix of sources like:
- Books, academic papers, and research journals
- Publicly available web pages
- Licensed datasets and structured knowledge bases
Unlike search engines, LLMs don’t have live internet access (unless explicitly designed for it, like Gemini). Their responses are based on their last training update, meaning they won’t always reflect real-time events. For example, OpenAI’s documentation on GPT-4 explains how its training data is periodically updated, and Google’s Gemini outlines its approach to real-time access.
Choosing the Right LLM for Your Needs
| Model | Strengths | Best For | Provider |
|---|---|---|---|
| ChatGPT (GPT-4) | Conversational AI, content generation, coding help | Blog writing, brainstorming, customer support | OpenAI |
| Gemini (formerly Bard) | Real-time web access, fact-checking | Research, verifying information, creative ideation | |
| Claude | Safety-focused, detailed responses | Enterprise use, legal and compliance support | Anthropic |
| DeepSeek | Multilingual, research-focused | Technical document analysis, AI-enhanced research | DeepSeek AI |
| LLaMA (Meta AI) | Open-source, highly customizable | AI model fine-tuning, research projects | Meta |
How to Use LLMs—Even If You’re Not Technical
You don’t need a Ph.D. in data science to get value from AI. Here’s how it works in three simple steps:
- Give it a prompt: Ask a question, make a request, or provide an instruction.
- Let AI process your input: The model analyzes patterns in its training data to craft a response.
- Get your output: AI generates text, summaries, ideas, or even code based on your input.
Real-World Example
If someone sends you a 60-page presentation, you can drop it into an AI tool and ask for a quick 10-bullet summary of the key takeaways. It’s still a good idea to skim through it yourself, but this way, you get the gist in under three minutes.
AI Isn’t the Future, It’s Now
AI is reshaping the way businesses operate today. A recent McKinsey report highlights how AI adoption is accelerating across industries, helping companies boost efficiency and unlock new revenue streams. Its helping companies automate workflows, create content faster and gain deeper insights from their data.
Tucker Borgman 
Leave a comment