Jack Dorsey: Square, Cryptocurrency, and Artificial Intelligence | Lex Fridman Podcast #91

Square shifted from a distrustful, credit-score–based model to a “trust then verify” approach, using transaction-level risk modeling to move approval rates from ~30–40% to 99%, dramatically expanding who can participate in the payments system.

From its first year, Square invested heavily in data science and machine learning to model risk and convince banks and card networks that a radically more inclusive onboarding strategy could still be safe and reliable.

Dorsey argues that fragmented banking partnerships and regulations slow innovation; a broadly adopted digital currency like Bitcoin could let companies launch products globally with a unified monetary layer, enabling faster, more equitable economic participation.

He’s less focused on sci‑fi existential threats and more on opaque models making high-stakes decisions (lending, health, content) and the growing gap between deepfake/identity creation technologies and our lagging ability to detect and counter them.

Citing truck drivers and cashiers, Dorsey agrees with Andrew Yang that AI and automation will remove vast numbers of jobs quickly, and that people need a financial floor—via universal basic income—to survive while they retrain and transition.

Building on Yuval Harari’s ideas, he notes that recommendation systems can know individuals’ preferences better than they do, and sees meditation and self-inquiry as tools to preserve awareness that we are delegating choices to algorithms.

Dorsey uses practices like one-meal-a-day fasting and public speaking challenges as experiments to push beyond comfort; he finds these constraints improve mental clarity, expose personal limits, and help him prioritize what truly matters.