AI Crash Course from Caleb
A two-minute read to get you up to speed on AI.
“Okay so WTF is an AI?! and why is it all I’m hearing about?”
Crash Course
AI = Artificial intelligence.
And it’s been around for a lot longer than you think.
Like since the 50s…
Ever since we decided it would be useful to program computers to move from basic logic to more advanced reasoning.
When most people talk about AI today, what they’re usually referring to is a Neural Network.
Neural Network = replicating the advanced cognition of a human brain with a computer.
“Okay great, so why won’t anyone shut up about it?”
The Breakthrough
Because neural networks are getting smarter. A lot smarter.
The latest and greatest advancement is something called a GPT.
GPT = Generative Pre-Trained Transformer.
(remembering that should impress your coworkers at lunch)
Quick Breakdown
Each of these words is an important feature in what makes GPTs so useful.
Generative = original responses
Pre-trained = knowledgeable on lots of information
Transformer = better understanding of what is being said to it
There are different types of GPTs, and these are called models.
Popular models right now include; ChatGPT, Claude, Gemini and Grok
(hopefully some of these you would of heard of by now)
Each a different company’s take on the same technology.
Behind the Math
Here’s usually where you get the big, over-complicated explanation on how AI actually works…
But remember there are two reasons why people make AI sound complicated.
They don’t understand it enough to explain it simply.
OR They just don’t want you to understand (they choose to gatekeep)
We don’t gatekeep here.
While the name Generative Pre-Trained Transformer can be a bit of a mouthful… this is essentially what it does…
Ready?
A program that can reason like a human…
Yep.
That’s it.
GPTs can take user information and, using some very clever math, understand what’s being said or shown and respond in kind.
Beyond Understanding
Now let’s take this into business.
“Great so it can read my emails?”
“Is that it?”
Well understanding input is only the first step…
Here’s the next step…
Models **become extremely valuable when they are empowered with actual utilities.
AKA being able to interact with other programs.
(you’ll commonly hear these called tools or connectors, you might even hear people talking about APIs)
If the model is the brain, then the tools are the limbs.
So now not only does it read your email but now… it can reply.
“Okay so it can do my emails.”
“What else can it do?”
Finally we’re asking the right questions.
Where To Start?
The truth is at this stage it seems like the possibilities are endless.
And with infinite possibilities it can be hard to know which direction to move in.
A good question to ask yourself is:
“How much of my business’s inputs and outputs are digital?”
The answer for a lot of businesses we work with is surprisingly quite a lot.
And it gives us a great idea of just how much can be automated right away.
Time To Talk To Us
The goal of business will always be to solve a problem and make money doing so.
Which means one of two things:
Increasing revenue
Cutting costs
AKA - moving the needle.
Now how we do that using AI depends entirely on your business.
Because every business is unique.
Unique leadership, unique staff, unique culture, unique SOPs.
Meaning the AI solution for your business needs to be unique.