What does it actually take to go all in on becoming an AI-native company. Not as a startup, but at over a billion dollars in revenue? Why are agents "forgetful", and what does that mean for how you organize your company's knowledge? And who really captures the value in enterprise AI: the model providers or the routing layer in between?
In this episode of AI to the DNA, the podcast for everyone who wants to really understand and apply AI, I sit down with Aaron Levie. Aaron is Co-Founder and CEO of Box, the enterprise content management platform he has led for over 20 years – making him one of the longest-serving CEOs in Silicon Valley. Today, he's rebuilding Box from the ground up as an AI-native company.
We talk about why simply dropping agents into existing processes rarely delivers the expected upside, what big companies can learn from startups that treat agents as first-class citizens, and why knowledge management is becoming one of the core raw ingredients of this new industrial revolution.
We also discuss the practical side of the transformation: why Box has had to change its engineering toolkit three or four times in just two years from GitHub Copilot to Cursor to Claude Code and Codex, what forward deployed engineers actually do at customers, and why the ROI of AI has to be owned by business leaders, not the CIO.
From there we zoom out: Will there be one agent to rule them all, or 20, 30, 50 agentic systems per enterprise? What happens to jobs when companies use AI to get more output per person instead of cutting headcount? And why does Aaron remain optimistic about the future of work after two decades of building Box?
I'm Christoph Magnussen and at AI to the DNA, I'm not just asking how to use AI. I’m asking why it works the way it does. This podcast is for everyone who wants to move beyond tool-tourism and master the real craft of AI: Learning to think, work, and build side by side with intelligent systems!
☀️ 🎓 Die Blackboat AI Summer School: Vier Wochen Online-Sprint mit Hands-on-Aufgaben, um vom Kl-Ausprobieren zu echten Pro-Skills zu kommen. Vom Tool-Tourist zu eigenen Agent-Systemen mit unserer AI Summer School!
Spare jetzt 250€ mit dem Gutscheincode: ATTD-SUMMER
-> Jetzt anmelden auf https://academy.blackboat.com
⭐️ Wenn dir AI to the DNA gefällt, freuen wir uns über eine 5-Sterne-Bewertung. Damit hilfst du uns enorm, den Podcast sichtbarer zu machen und noch mehr Menschen mit unseren Themen zu erreichen. Danke für deinen Support!
- (00:00) - Intro
- (01:33) - Organizing Knowledge for Agents: The Key Question of the Agentic Era
- (04:10) - Who Gets It Right? Startups with Agents as First-Class Citizens
- (06:50) - All In on AI Native: How Box Rewires Every Process
- (10:06) - Fragmented Data: Why Enterprises Lack Sources of Truth
- (12:56) - Forward Deployed Engineers: Heavy Human Work That Can't Be Automated
- (14:40) - New Models, New Evals: The IT Organization of the Future
- (16:14) - Locked Models vs. the Routing Layer: Who Captures the Value
- (21:09) - Budgets and ROI: Why Business Owners Must Own AI, Not the CIO
- (25:31) - Security vs. Speed: Changing the Engineering Toolkit Four Times in Two Years
- (28:39) - Headless or GUI: How Agents and Interfaces Work Hand in Hand
- (31:18) - Aaron's Personal Agent Stack: Claude, Codex, Cursor & Perplexity
- (32:29) - Not One Agent to Rule Them All: 20, 30, 50 Systems per Enterprise
- (34:40) - "We Have Copilot, So We're Doing AI": Talent, Competition, Catching Up
- (37:38) - Software of One: More Software, Not Less
- (40:49) - Customer Relationships as the Sustaining Moat
- (42:49) - Rebuilding Box AI Native and 20 Years of Drive
- (45:23) - An Optimistic Case for the Future of Jobs