After years of building GenAI solutions in the bustling tech corridors of Mumbai, I’ve arrived at Syracuse University with a singular mission. It’s a mission born from a fundamental shift I’ve witnessed not just in my own work, but in the very fabric of software itself. This isn’t just a career pivot, it’s a commitment to designing the human interface for the next generation of intelligence.
🌍 From Software 2.0 to 3.0: Why I Chose This Path
My journey began with curiosity and code. Working on GenAI deployments for global clients taught me the incredible power of Software 2.0, the world of neural networks and machine-learned systems. But it also revealed the limits of purely algorithmic intelligence.
We are now entering the era of Software 3.0, where we program massive neural networks not with Python or C++, but with natural language. English is the new programming language. This paradigm shift means the next great leap in AI won’t be about faster models, but about more human ones. That realization brought me here.
Human-Centered AI, to me, is the essential layer we must build on top of this new software paradigm. It’s the art of embedding emotional intelligence, ethical reasoning, and collaborative fluency into systems that can now understand our words. Collaborative robotics is its physical manifestation: machines that don’t just automate tasks, but use this new intelligence to co-create experiences with us.
🧠 My Focus: Building the Human-AI Partnership
Karpathy describes today’s LLMs as having a fascinating but flawed “psychology”, possessing encyclopedic knowledge alongside “jagged intelligence” and a tendency to hallucinate. My work aims to address this challenge head-on. If the AI is our new collaborator, we must design it to be a better partner.
I’ve officially launched my M.S. in Human-Centered Artificial Intelligence at Syracuse University, with a focus on:
- Designing for Software 3.0: Building ethical and empathetic architectures for Large Language Models.
- True Human-Robot Collaboration: Moving beyond simple commands to create seamless human-robot partnerships based on shared context and understanding.
- Designing Eve v1: My capstone project to build a modular AI assistant with memory buffers, emotional awareness, and bias filtration, an attempt to solve the “anterograde amnesia” Karpathy notes in current models.
✍️ First Steps on a New Foundation
- Initiated a weekly reflection series to document learnings, challenges, and breakthroughs
- Began sketching Eve v1’s emotional logic inspired by rituals and symbolic design
- Joined student networks and research clubs to integrate culture and collaboration into my academic rhythm
🔭 The Vision Ahead
In five years, I aim to architect AI companions that don’t just respond, but resonate. Systems that understand context, adapt to emotion, and co-evolve with their users. I want to contribute frameworks that bridge technical rigor with human nuance, embraced by both research labs and real-world communities.
This journey is not just about building tech. It’s about building trust, meaning, and shared intelligence.
🌀 Whether you’re a fellow technologist, a curious mind, or someone exploring the soul of machines, thank you for being part of this journey. I’ll be sharing weekly reflections, project updates, and design insights right here.
Let’s build something that feels. Let’s build something that matters.
