`Welcome to` Lossfunk

We are an independent AI research lab founded by Paras Chopra, focused on advancing the frontiers of machine learning, large language models, and intelligent reasoning. We’re a playground for deep thinking on artificial intelligence, its possibilities, risks, and radical potential. We believe in the power of focused, independent inquiry. Our mission is simple but ambitious:
Push the boundaries of AI reasoning through novel, high-quality research.

On this Substack, you’ll get:

Original AI research and technical breakdowns
Musings on what it means to reason, learn, and adapt
Behind-the-scenes looks at how research ideas evolve
Updates from our residency & initialise program, meet the people shaping the future of AI
Occasional thoughts on the philosophy of intelligence and learning

Whether you're an ML researcher, engineer, or simply fascinated by intelligence, this space is for you. Think of this as our open lab notebook and invitation.

Why This Exists?

AI is reshaping the world. But who gets to shape AI?

Lossfunk was created so independent thinkers can challenge norms, break models (sometimes literally), and publish meaningful work with no institutional red tape. We support researchers through funding, mentorship, and community. And this newsletter is where we share the journey.

Who’s Behind This?

I'm Paras Chopra — a lifelong technologist, founder of Wingify (makers of VWO), and AI enthusiast. I’ve been exploring neural networks since writing my first one in Visual Basic 20 years ago.

Lossfunk is my way of supporting open, impactful AI research.