Why I invested

Personal AI is tackling one of the most fundamental human challenges: memory. We forget about 80% of information within three days, and our knowledge remains locked in our minds, inaccessible to others. The founding team (Suman Kanuganti, Sharon Zhang, and Kristie Kaiser) have a clear vision: build an AI for each individual to augment human biological memory.

What drew me in was their quote: "You don't solve the human, you solve the problem." They're not trying to enhance brains surgically or chemically. They're building tools that solve the memory problem externally.

The thesis

Personal AI represents a fundamental shift from centralized, corporate-owned AI to personalized, user-owned models. Their thesis rests on three pillars:

1. Decentralization of ownership - Self-sovereignty of data through Web 3.0 principles, giving individuals control over their identity and personal AI
2. Decentralization of AI - Personal Language Models (PLMs) trained on individual data, eliminating the bias inherent in aggregated models
3. Decentralization of economy - Tokenization frameworks that let individuals monetize their own AIs for their economic benefit

What excites me

1. Personal Language Models - Each user gets their own 120M parameter model trained entirely on their memories, not a shared LLM
2. User ownership - Your Personal AI is owned and managed by you, never sold
3. Proven retention - 35% Day 30 retention for users with 100+ memories, 52% for 1,000+ memories—above a16z's "Great" benchmark
4. Capital efficiency - 80%+ profit margin per subscriber, up to 40x more cost effective than LLM-based systems
5. Founding team - Four successful exits totaling over $20B combined

The future of AI is personal.