How to enable the new ChatGPT: Advice from OpenAI

OpenAI's latest ChatGPT model, OpenAI o1, is a true powerhouse of reasoning. This new version doesn't just spit out answers. It pauses, thinks, and reflects before responding. It's like chatting with an attentive friend who carefully considers your words.

According to OpenAI, this internal thought process leads to more considered answers. For founders and entrepreneurs using AI in their companies, ChatGPT o1 is a game changer.

How to enable the new ChatGPT: Tips from OpenAI

Keep the prompts simple

What you learned about writing in school is becoming less and less relevant. To achieve optimal results online, formal and complicated is not the way to go, but relaxed and clear. The same goes for the new ChatGPT, which focuses on clarity. Avoid long-winded instructions and keep it short.

Instead of “Can you come up with a list of 20 possible product names for a new eco-friendly water bottle, taking into account factors like environmental impact, target audience, and current market trends?”, try “5 catchy names for an eco water bottle.” The model figures out what you're looking for and fills in the gaps. Like a human, it will grasp the context and read between the lines.

Let it think for itself

When a new team member joins your company, teach them to follow your processes. When an experienced employee takes the reins, expect them to have their own. ChatGPT wants to be the experienced employee who has free reign over your tasks and uses their own methods to find the right answer.

This means there are no more Chain of Thought (CoT) prompts. You don't need to tell it to “think step by step” or “explain your reasoning.” ChatGPT is already prepared for this. The new ChatGPT does its own mental gymnastics and answers may take longer while it thinks. Just ask the question and let it go.

Divide your instructions into

When you give advice to someone, pause as you speak. When you give instructions in emails, separate them with paragraph breaks and line breaks. The same applies here.

The new ChatGPT wants you to use separators. These include triple quotes, XML tags, or simple section titles – they are like signposts for the AI. Using them helps ChatGPT understand which part is which and avoids confusion.

You have many options for separators: Markdown headings, hyphens, asterisks, numbered or labeled sections. Break up your prompts, break up your examples, and make sure your instructions don't get jumbled.

Do not overload the chat

If you give a person too much information, they will shut themselves off. Your hairdresser doesn't need your life story, just how you want your hair done. Your accountant doesn't need every detail of your career, just what is relevant to your accounting today. Think similarly here.

RAG, or Retrieval-Augmented Generation, means you feed additional information into ChatGPT. But with the new model, less is more.

OpenAI warns that overloading ChatGPT with data can lead to overthinking and slower responses. Where once you might have used pages of examples for a simple task, now you should only include the most relevant parts.

How to request a new ChatGPT for optimal results

These changes mean you need to rethink how you use ChatGPT. It's no longer about crafting the perfect, long prompt. It's about asking precise, targeted questions. By customizing your prompt style, you'll get faster, more accurate answers.

OpenAI emphasizes that the key to success with o1 is to trust the model's ability to understand the context and generate relevant answers without excessive guidance.

When technology evolves so quickly, you need to keep learning. Make the most of the new ChatGPT with these four simple steps.