Not a day goes by when some high-end organization or educational institution releases the next big thing in AI. From text-to-video tools to AI-powered research assistants, the pace of innovation is relentless. It’s getting harder to keep up, but also more exciting. Every new release has the potential to reshape how we work, learn, and create.
People at Stanford University have come up with an innovative idea that revolves around creating Wikipedia-style articles. STORM is an AI system that can write Wikipedia-style articles from scratch by searching the web for information. Its upgraded version, Co-STORM, takes things further by letting humans work alongside the AI, making the whole research and content curation process more in tune with what people actually want.
It’s not perfect, though. What it writes still needs a lot of polishing before it's ready to be published. But seasoned Wikipedia editors say it’s a useful tool to kickstart their writing process and gather relevant info quickly.
Now you know what this tutorial is going to be about. We will show you how to use Storm AI to create Wikipedia-style articles for topics that you are passionate about. You will learn how to access Storm, provide a prompt, ask for edits, and finally download the article for an in-depth review.
By the end of this tutorial, you’ll be able to:
- Access the Co-Storm via the website
- Write a custom prompt to initiate the article writing process
- Use Co-storm to interact with Moderators and human users
Let’s jump right into it!
Step 1 - Access the Co-Storm via the website
Since this is a new website, you might need to create an account. Access Storm and create an account by clicking ‘Get Started.’

Use a Gmail account to sign in to an account. Agree to the terms of service. Click ‘Continue.’

Agreeing to provide the details of your interaction with Storm is necessary. Give your consent and click ‘Finish.’

You’ll be welcomed to a strikingly familiar interface.

The prompt box gives you the option to select between Storm and co-storm models. Co-storm is a collaborative tool that allows you to interact with LLM experts, moderators, and human users. You can get more information about the article, get citations for certain passages, or just interact with them to get questions related to your article. And there’s another model that is focused on Data streams. It’s called Data-storm. This model lets you conduct deep research with data for a specific topic.

Another feature is ‘Discover’ in the left navigation pane. Here you can discover a wide range of topics posted by other users. You can get inspiration for your prompt by viewing any of those articles.

That’s it for the interface. Let’s move on to create our first prompt with Storm.
Step 2 - Write a custom prompt to initiate the article writing process
Generating a Wikipedia-style article is a process rather than a single output. Storm first gathers necessary sources and lists them so you can review them. Then, it gears up to write an article based on the topic you suggested in your prompt. This process is crucial in writing a factual article.
For example, you want to write an article about how male orca whales behave around sharks and other large marine mammals. Use the following prompt to test Storm capabilities.
Prompt:
Describe how male orca whales interact with marine mammals, including hunting, play, aggression, and differences across regions and ecotypes.
Important note: Storm doesn’t support prompts that are more than 20 words. Keep your prompt short so it won’t exceed 20 words.

That looks fantastic. There is no option to ask follow up questions to add more topics to this article. Storm generated the article in one-go. Now you must initiate a new chat and ask about additional information. Unfortunately, at the time of writing this tutorial, you might have to copy/paste the entire thing yourself rather than get a PDF of all the entire article it generated.
Step 3 - Use Co-storm to interact with Moderators and human users
The Co-storm feature is ambiguous. The nature of the conversations remain a mystery. We have no idea whether the participants are actual humans or just intelligent bots. We will explore this feature together and see whether it’s has any good use for the article we wrote in this tutorial.
Right now, it seems you have to provide a prompt that generates an article with co-storm. Then a group of ambiguous people start a conversation about it and that might lead to an improvement or addition of certain topics in the article.
Initiate a new chat and select ‘Co-storm’ from the drop-down menu.

Let’s use a fresh prompt for Co-storm.
Prompt:
Describe how orca whales give birth, including gestation, birthing process, maternal behavior, and care for the newborn calf.

It seems there are various AI models that pose as experts and moderators and ask relevant questions based on the prompt. Users can comment by asking questions or adding information.

That’s it for this tutorial, AI knowledge-seekers. Storm is still in its infancy. It’s a developing tool. Who knows, they might create something that helps researchers find relevant information or get new dimensions on their topics they want to research. n
0 Comments