• Member
• September 11, 2025

The VLSI AI Challenge

1. The Srijan Sanchar Vision for AI

This competition is a call to action for the next generation of AI innovators. It is built on Srijan Sanchar's belief that true breakthroughs in artificial intelligence will not come from isolated labs or linear processes, but from the collaborative wisdom of a networked world. By leveraging the principles of Very Large Scale Innovation (VLSI), we aim to transcend conventional methods and discover novel AI applications by connecting the dots between disparate fields, disciplines, and ideas. This challenge is not just about writing code; it's about pioneering a new model for AI creation by embracing the "Extended Enterprise."

2. Competition Themes: The Extended Enterprise in Action

Participants will choose one of the following challenges, each designed to embody the principles of VLSI and Extended Enterprise Innovation.

• Track 1: Bridging Disparate Expertise

  • Description: An AI problem that requires integrating knowledge from two or more seemingly unrelated domains (e.g., fluid dynamics and biological nervous systems).

  • Challenge: Develop an AI that can solve a complex manufacturing problem by applying the principles of a non-manufacturing discipline. For instance, an AI for optimizing a supply chain based on the chaotic and random behaviors of a natural ecosystem. This challenge requires participants to act as "connectors," identifying and merging "unknown resources" (concepts from different fields) to create a new solution.

  • Metrics: The elegance of the interdisciplinary connection, and the unexpected nature of the solution.

• Track 2: The Black Box of Collaboration

  • Description: Participants will be given a "black box" AI system, developed in a different field, and will need to adapt it for a new purpose.

  • Challenge: Take an AI designed to optimize a retail inventory and re-purpose it to manage a city's public transit system. Participants must not have access to the original source code or training data. They will need to collaborate with a "simulated" external partner (providing limited, curated data) to infer the system's logic and adapt it. This simulates the real-world challenge of leveraging an external startup's pre-developed solution without full transparency, a core tenet of Extended Enterprise Innovation.

  • Metrics: The efficiency of the re-purposing process and the performance of the adapted AI in its new domain.

• Track 3: The Holonomic Dual

  • Description: This challenge is inspired by the "holonomic duals-based method of problem-solving." It focuses on finding a solution by inverting the problem's primary constraint.

  • Challenge: Instead of creating an AI that finds the most efficient route, create an AI that finds the most inefficient route, and then use the inverse of that logic to discover an entirely new, non-obvious optimal path. This non-probabilistic approach avoids linear thinking and reveals new solution spaces.

  • Metrics: The novelty and effectiveness of the "dual" solution, and the clarity of the underlying methodology.

3. Evaluation Criteria

Submissions will be evaluated on the following:

• VLSI Alignment (40%): How effectively does the solution demonstrate the principles of connecting disparate ideas and leveraging an extended network of knowledge?

• Innovation & Creativity (30%): How novel and creative is the core idea? Does it offer a new way of thinking about the problem?

• Practicality (20%): Can the solution be practically implemented and scaled?

• Methodology & Presentation (10%): The clarity and depth of the explanation and documentation.