OpenAI says Prompt Masters are ahead of the game


Hey Reader,

So last week, OpenAI has published their 5 levels towards AGI, and they say that we are at Level 1, and GPT-5 will be the first Level 2.

Some of our students and clients are already on Level 3 though.

How could that be possible?

Here is how OpenAI imagines our AI future:

Level 1: Chatbots, AI with conversational language

Level 2: Reasoners, human-level problem solving

Level 3: Agents, systems that can take actions

Level 4: Innovators, AI that can aid in invention

Level 5: Organizations, AI that can do the work of an organization

Now, even though these definitions are quite vague, there is a very important point: these are only true without extra tools.

But it is totally possible to give the AIs these extra tools that can make them Level up on this ladder.

AI Agents are systems that can take actions.

ChatGPT can't do this, because there you only chat with the model and nothing happens outside of that interface.

But by connecting it to your own systems, suddenly it can take actions for you.

Here's a simplified example:

Let's say a New Email comes in, and we configured an AI agent to handle it.

The first GPT module in this agent acts as a moderator. It basically decides:

  1. What needs to happen next
  2. Who needs to do it
  3. Any extra instructions needed

This GPT will respond with only one word, like "email" "slack" "notion", and then only that route will be triggered.

So if the moderator decides that this is an email the other agent can easily respond to (based on certain critera) it sends it to that.

If it's something that the team should be aware about, it sends a Slack message.

And if it's an update of a client project, it creates a new database item in Notion.

Now, this is an oversimplified example for visualization purposes, real AI agents are a lot more complicated than this one.

They are usually made up of multiple Make scenarios and dozens of modules within each of them.

So something more like this:

This agent was created by one of our Prompt Masters, Janos, who has created many of these customized AI agents that fit our clients businesses like a custom tailored suit.

Do you want us to create agents for you as well and save your time and money?

If yes, book a quick 15-minute call with our CEO David, and let's see if we can create what you want. If you don't know what you want, you should also book a call, because David can also help you figure that out.

Think faster,

Dave Talas, co-founder of Promptmaster

Promptmaster Newsletter

Sign up if you are overwhelmed with all these AI tools out there, and want a clear path on your AI journey.

Read more from Promptmaster Newsletter

Hey Reader, After writing my Rise of the Phoenix email yesterday, I’ve copy-pasted the whole thing to ChatGPT, without any prompt. Based on what it said back to me, I’d like to reflect on how it can hurt your confidence if you’re not careful. So basically it said: “Great email, very heartfelt and emotional yet still grounded. Here are 6 ways you can improve it and make it better. [List 1-6 with specific examples]” My next prompt was: Yeah, that’s what I thought! 😃 But imagine if someone is...

Hey Reader, It’s been a while since I’ve last written to you. You might have noticed that this email is coming from a new domain (if it landed in your promotions, please move it to Primary, and add me as a Contact). I’m leaving the Promptmaster brand behind, there were so many problems with the business model, and so much baggage from previous founders that I couldn’t keep the company profitable, and I’ve realized I wanted to do something else, and do it differently. I’ve had crazy successes...

Read time: 3 minutes Hey Reader, It is one thing to know the potential of AI Language models like ChatGPT or Claude, but it's also important to know their limitations. In this email, I want to show you 5 things you should avoid when using ChatGPT (or Claude or Gemini or LLama, etc.) 1/ Treat it like a know-it-all search engine ChatGPT is not a search engine. It is simply a very expensive predictive keyboard. All it does is it just guesses what's the best next word to say in a scenario. It...