In the ever-evolving landscape of artificial intelligence, Stanford University’s STORM has emerged as a powerful new player, promising to revolutionize the way we research and synthesize information. This open-source project, capable of generating comprehensive reports on any given topic, complete with sources, raises intriguing questions about the future of knowledge curation and dissemination.
As STORM aims to bridge the gap between vast online information and individual comprehension, it’s worth examining both its potential benefits and the challenges it may present to traditional research methods and critical thinking skills. With the ability to produce Wikipedia-like articles from scratch, STORM represents a significant leap forward in AI-assisted content creation, but at what cost to human intellectual engagement?
The site is extremely slow, which for a free generative AI tool from a university that is running on “donated Azure credits” is to be expected. But there is more. When entering a prompt to “Create a New Article” it would ask you for clarifying information, even if you used the same prompts that were provided in the Discover/Examples section of their site.
I found this confusing because the prompt was pretty self-explanatory. For example, one of the prompts I tried was “Explain to me what STORM is and how it compares to existing generative AI tools.” The response I got was to elaborate and when I did it threw an error. It took so long to get the system to finally register something, that by the time it worked the topic and content I was looking for initially was no longer the same.
When it does work though, it is pretty amazing. Not only does it show you all of the sources as it goes through them, but it also puts together a very thorough article, with cohesive content, backed by sources that actually resolve when checked. I wanted to show you how this looked but the service is temporarily disabled!
If this was around during my academic career, I would 100% have used it, the wait times would be worth it. But as the screenshot below shows, it doesn’t always work. I think this is a great example of how important infrastructure, scalability, caching, and DevOps are going to be instrumental for any AI company to succeed.
The date says December 31, 1 at 7:03 PM, but this occurred at 11 AM on September 24, smh
STORM shows the promise and peril of academic AI tools: brilliant research capabilities hampered by infrastructure reality checks.
Who’s walking who?
I was walking my dog the other day and he was pulling me all different directions, and an old gentleman said “who’s walking who?” we both chuckled. But I am worried that AI is doing the same thing to humans.
Before you endeavor on a task with AI, you should write down what you are trying to do. Have that be your North Star and try to stick to it, don’t let Midjourney, STORM, or whatever tool decide which direction you are going in. AI should be a tool that helps enhance your creativity or initiative, not drive it.
The beauty of STORM lies in its transparency—showing you the sources it’s consulting in real-time, much like watching a researcher’s thought process unfold. This visibility sets it apart from other AI tools that operate as black boxes. When it works, it produces genuinely impressive, well-sourced content that could serve as an excellent starting point for deeper research.
However, the technical challenges highlight a broader issue in the AI space: the gap between research prototypes and production-ready tools. Stanford’s STORM represents cutting-edge research, but the user experience reminds us that innovation and reliability don’t always develop at the same pace.
AI should enhance your creativity, not drive it. Write down your goal before you start—that's your North Star in the AI wilderness.
The real question isn’t whether STORM can replace human researchers, but whether it can make us better researchers. The tool’s ability to synthesize information from multiple sources and present it coherently is undeniably powerful. Yet the frustrating user experience serves as a reminder that we’re still in the early days of AI-assisted research.
As these tools mature, we must remain mindful of maintaining our intellectual agency. The goal should be to use AI to amplify our capabilities, not to let it determine our direction. STORM, when it works, offers a glimpse of that future—one where AI handles the heavy lifting of information gathering while humans focus on analysis, synthesis, and creative thinking.
If you have some time… a lot of time and want to check STORM out, you can visit: https://storm.genie.stanford.edu/
Just remember to keep a firm grip on that leash.
Vatché
Tinker, Thinker, AI Builder. Writing helps me formulate my thoughts and opinions on various topics. This blog's focus is AI and emerging tech, but may stray from time to time into philosophy and ethics.
