"Quality over quantity. It is never about the number."

Albert Chun is the Founder of AI Circle — a vetted community for business leaders, founders, and operators building at the frontier of AI. Before AI Circle, Albert was a teacher in two of the country's poorest congressional districts, a Harvard grad student, an insurtech operator who scaled from 3 reports to managing 211 people across 13 M&As, and a senior leader at Invisible Technologies — where his team helped train what became the #1 open source frontier model in the world.

In this episode, Dennis and Albert get into what it actually looks like inside an AI company building on the frontier, why the AI ecosystem is over-indexed on credentials and under-indexed on experience, and the deeper human problem AI Circle is really solving.

In this conversation, we cover:

Why Albert walked away from C-level offers at triple the pay to take an AI role that "felt like the A+ problem in the A+ space of the industry"
What training a frontier model actually looks like day-to-day — and why reading one arxiv paper can put you ahead of almost everyone
How Albert thinks about AI ROI when model outputs are stochastic, loss functions don't map cleanly to revenue, and causation is a finger in the wind
The gatekeeping problem in AI — why polysyllabic jargon like "chain of thought reasoning" is really just "me sharing what I thought and why I did it"
Why Andrew Ng and Andrej Karpathy are both building education — and Albert's theory that education, not GPUs, is the biggest bottleneck in AI acceleration
Writing in public since Xanga and Friendster, the story of a broken pair of glasses that led to a $250K donation, and why corporate "stay in-house" advice was the wrong tea leaf
Why AI Circle is bootstrapped, not VC-backed — and the fight story from sophomore year that shaped how Albert thinks about community size forever
The deeper problem AI Circle is really solving: community, connection, and relationship as life gets harder, and friend circles get smaller

If you are a business leader trying to separate real AI signal from buzzword noise — or a founder thinking about how to build trust-based community in an era of vanity metrics — this is one of the most honest conversations on the pod.