Inspiration
As automation and AI systems increasingly make decisions in hiring, healthcare, finance and governance, responsibility often becomes unclear. When something goes wrong, people say, “the system decided.” But systems don’t create themselves, humans do.
I was inspired by a simple but powerful question: Who made the last human decision before automation began?
LAST HUMAN INPUT was created to preserve that moment, the point where human intention becomes machine action.
What it does
LAST HUMAN INPUT is a human-accountability system that records, preserves and displays the final human decision made before a process becomes automated.
The system allows users to log: the decision made the goal behind it assumptions acknowledged risks the responsible human (optional) It then uses AI to generate a neutral summary explaining: what the human intended what they believed what risks existed
Each decision becomes a permanent record that shows the origin of automation and reminds future users that every system began with a human choice.
How I built it
The system was built as a lightweight, transparent web application using: HTML for structure CSS for minimal, calm design JavaScript for interaction PHP for backend logic MySQL for storing decision records Gemini AI API for analyzing human intent The architecture follows a simple flow:
User → Form → PHP Backend → Database → AI Analysis → Stored Record → Display
The design intentionally avoids complexity so the concept remains clear, accessible and explainable.
Challenges I ran into
One of the biggest challenges was balancing technical functionality with philosophical clarity. It was tempting to add advanced features, dashboards, or analytics, but doing so risked distracting from the core message.
Another challenge was designing AI output that remains neutral and reflective rather than judgmental or predictive. The system needed AI to act as a mirror, not a decision-maker.
Accomplishments that I'm proud of
Creating a system that feels meaningful rather than just functional Designing an experience that communicates seriousness and responsibility Building a concept that judges immediately understand Turning a philosophical question into a working prototype Most importantly, I’m proud that the project sparks reflection. People don’t just use it, they think about it.
What I learned
I learned that impactful technology is not always the most complex technology. Clarity of idea, purpose and design can matter more than feature count.
I also learned that AI can be used not only to automate decisions, but to help humans understand their own decisions more clearly.
What's next for LAST HUMAN INPUT
Future versions could expand into: AI governance and compliance tools Decision audit trails for organizations Policy transparency platforms Ethical review systems for automated technologies
Ultimately, LAST HUMAN INPUT could evolve into a global accountability layer that ensures no automated system exists without a traceable human origin.
Log in or sign up for Devpost to join the conversation.