Stacy Chin describes founders mass-producing AI-generated grant applications without understanding what they're doing. One PI submitted over 80 SBIR proposals in a year, all superficial fluff. Founders were never taught how to use AI effectively or what makes a strong proposal.
In fact, a PI submitted over 80 applications in a single year, all clearly AI-generated. And now that's wild. And I witnessed this myself, founders using AI to crank out entire grant applications in minutes. But frankly, if you read them, they're just pure fluff. So over the past year, I've had people come asking me for reviews. In 2 seconds, I can tell that text was AI-generated. No rigor, no depth, just filler content that looks impressive but says nothing. And honestly, this can be a waste of time for resources both on your end as a founder, and for reviewers at federal agencies behind the scenes. And here's the part founders don't realize: it's not their fault. Most founders were never taught how to use these tools properly, or in fact what reviewers are looking for in a great application. So no one has any idea what a strong proposal looks like. So when they plug their ideas into AI, they get something that sounds smart but completely misses the mark. And then I realized that gap. It just hit me. And no one has taught me these processes either of how to use AI the right way, how to build prompts that produces actual meaningful content, or even how to write strong SBIR applications.
Applying for NIH SBIR in Jan 5, 2027? You Should Be Starting NOW. Here's why...
How to Get Invited to Apply for $305K in NSF SBIR Funding || Step-by-Step Guide
