Hey,

Over the last few weeks, we've sent out a few different emails in this newsletter about AI agents.

We did that because we see that's the next big thing.

There are essentially 3 levels of using AI for work. (well, David says 5, but the last 2 are pretty far for most everyday users, so I simplified it)

Level 1: ChatGPT

Using ChatGPT can save hours every day. Especially if you know how to prompt it right, but even without Prompt Engineering skills, you can get decent outputs out of it.

Time saved: 1-2 hours per day

Level 2: Custom GPTs

Once you use ChatGPT repetitively for certain tasks, you can save the prompts in a Custom GPT, and save the time of rewriting the prompt. For each task you have, you can configure a custom GPT this way.

Maybe even add your own knowledge base and even configure external API actions. You can also call the GPTs with the @ sign in a ChatGPT chat thread, so you can save time there and have multiple Custom GPTs talking in the same conversation.

Time saved: 1-2 hours per day

But let's be honest. These will not magically make you 10x more effective. An hour saved here, 30 minutes there, etc. They add up, but it's not the real unlock.

Furthermore, this has a downside, that now you are doing almost 1.5x of work, creating 1.5x of total content (any data) which means your admin overhead goes up by 1.5x as well.

You can do more tasks per day, so you spend a tiny bit more time in the project management software, talking to your team, and handling all this extra information flux in your operations.

But it's still you or your team that has to open ChatGPT and prompt it do make it do things. You can't make ChatGPT run while you sleep or you're on a holiday.

And this is where AI agents come in

Level 3: No-code AI Agents

Before I get into why these are a big deal, let's break down what this phrase means.

Agent: A system that can sense it's environment and take action that affects the environment based on its conditioning. (for example, a thermostat is an agent: if room temperature falls below 20 degrees, turn on the heat ➔ it senses the environment and changes it through the heating system)

A.I: Artificial Intelligence - a computer system designed to do tasks that previously required human intelligence. In our case LLMs like GPT-4o are integrated into the agent.

No-code: All that created without writing a single line of code. So non-techie, non-programmers like you and me can create these agents now.

So here's why this is actually a big thing.

ANYONE with a computer and an internet connection can now create an AI agent for themselves, that will get better as more powerful LLMs are released.

And these AI agents can be triggered by literally anything.

For example:

• New email comes in
• New customer
• Change in Project Management database
• New slack message
• Meeting scheduled
• Meeting concluded
• New file in folder
• etc.

You get the idea.

Every single thing that happens digitally in a business, an AI agent can sense it, and do something about it. Automatically. Instantly. Better than humans.

The key is: how well can you configure it? Meaning, how good and reliable is your prompt? Does it give a useful response 90% or 99% of the time?

How well did you configure the triggers and filters?

There are a lot of tips and tricks I have to share with you, and I've recorded the 10th module in our Prompt Master AI course last week, will release it this week, that's all about configuring these no-code AI agents.

If you'd like to get access to our linear AI education course that brings you from zero to no-code AI agent builder in ~12 hours, I invite you to become a Prompt Master today! Click this link, and if you have ANY question about the course, reply to this email with the question (we're working on adding info to our landing page)

Talk soon,

Dave Talas

P.S: We currently offer a 30 minute bonus consultation call next to new course students, so join now and let's hop on a call soon!

‍