How to create and automate with AI with the chat interface and nothing else - part III

How to create and automate with AI with the chat interface and nothing else - part III

AI in its current state is a wonderful tool for automation, but like with any other new tool, you have to know how to use it safely and efficiently.

You can use AI to automate your everyday tasks, with no coding, no deep knowledge of tech magic behind, with a simple access to one of the main LLMs through their chat interfaces. However, using it without basic understanding of its limitations might lead to repeating mistakes many of us have already made, and wasting your time unnecessarily.

There are certain AI limitations you have to be aware of:

  • context window limitation,

  • bias to comply and flatter,

  • lack of strict rules to follow.

Let’s dive in, explore what it means, and how to overcome the challenges it brings. In the previous articles, we covered the context window limitation and bias to comply and flatter. Today, we will cover:

Lack of strict rules to follow

The LLM works contextually, assuming what the next token should be. There is no hard code that says how the precise data should be handled, what value goes with what item in your table etc. It is getting better in understanding the situations where it needs to be precise, but there is no guarantee yet. The same goes for rules you assume it gets, but there is no definition of them anywhere.

How to handle it?

Don’t assume it knows something you do. Before jumping in a difficult task, make sure it knows the basic guidelines - ask the AI to clarify the rules it will follow. Read it through, then agree with it and continue with your conversation. It will help with both the rules following, and the context limitation mentioned above.

Example: You want to create a series of graphics to go on your social media account. You know that square format works the best on Instagram, and you want them to follow your visual style - for example blue and black drawn pictures with a short headline. You feed the AI with examples. Then, instead of saying: “create another post like this for my Winter Sale of all water cups” ask it “Describe the visual style of these posts and rules you should follow when creating a new one for my brand”. This way, the AI will destill your visual style, creating its own guidelines to follow when creating a new post for you. If there are any imprecise rules, clarify them. If something is totally wrong, correct it. From now on, this will be your “point 0” as mentioned in the first article of this series.

We will dive in the wild waters of AI related topics again. To make sure you don’t miss it, you can subscribe for receiving our newsletter and receive a free PDF with advice on AI automation here.

#AI #automation #AIautomation #chatGPT #LLM

Don't want to miss out on new articles, insights and tips? Sign up for our newsletter, and get a free PDF with tips on how to handle AI automation!

Sign Up for Newsletter