The Twenty Minute VCAaron Levie: Everyone is Wrong; We'll Have More Developers in 5 Years
Harry Stebbings and Aaron Levie on aaron Levie on agents reshaping enterprise software, jobs, and budgets.
In this episode of The Twenty Minute VC, featuring Aaron Levie and Harry Stebbings, Aaron Levie: Everyone is Wrong; We'll Have More Developers in 5 Years explores aaron Levie on agents reshaping enterprise software, jobs, and budgets Levie argues the US–China AI dynamic is primarily a commercial and economic race, not an existential sprint where small timing advantages decide outcomes.
At a glance
WHAT IT’S REALLY ABOUT
Aaron Levie on agents reshaping enterprise software, jobs, and budgets
- Levie argues the US–China AI dynamic is primarily a commercial and economic race, not an existential sprint where small timing advantages decide outcomes.
- He contends AI won’t eliminate most jobs because humans remain in the loop—AI shifts where review and accountability occur and reveals new bottlenecks that require more labor.
- A new “agent operator” role will emerge to redesign regulated enterprise workflows for agents, connect fragmented data systems, and manage ongoing model-driven workflow breakage.
- Software value will increasingly migrate from button-heavy UIs to robust APIs and embedded business logic, making “headless” platforms and strong governance layers essential.
- Agent adoption will trigger a cybersecurity surge and a major budgeting shift as token/compute spend moves from IT budgets into broader operating expense tied to business outcomes.
IDEAS WORTH REMEMBERING
5 ideasAI changes work review points more than it removes humans.
Levie’s core claim is that organizations still need human validation for regulatory, legal, and reputational reasons; AI compresses production time but does not remove accountability, so humans re-enter at higher-level review stages.
Developer demand will expand beyond tech into the remaining 85% of the economy.
He argues most non-tech industries (manufacturing, pharma, banking, agriculture) are under-engineered relative to their automation needs; coding agents make it feasible for them to “buy” engineering capacity they couldn’t access before.
Expect more output—and more constraints—creating new hiring needs (even in law).
By making it easy to generate contracts, memos, and filings, AI increases throughput, but courts, regulators, and approval processes remain slow; this can increase demand for credentialed reviewers (e.g., more lawyers) even if junior tasks change.
The next breakout enterprise role is an ‘agent operator’ who redesigns workflows.
This role blends technical skills (tooling, MCP/CLIs, prompt/workflow configuration) with change management, data readiness, and process redesign—especially in regulated enterprises where greenfield startup playbooks fail.
SaaS won’t automatically become ‘valueless databases’—business logic and governance matter.
Levie agrees some UI-heavy tools will see value shift toward APIs, but argues many systems (ERP, compliance-heavy content platforms) embed proprietary logic, permissions, auditability, and regulatory features that remain defensible in an agentic world.
WORDS WORTH SAVING
5 quotesWe haven't removed humans from the loop. We've just changed where they enter the loop.
— Aaron Levie
There are going to be more lawyers in the next five years than we have today.
— Aaron Levie
The workflow needs to be redesigned for agents, not for people.
— Aaron Levie
Agents are the solution to the problem that agents have caused.
— Aaron Levie
The budget of tokens will have to move out of IT spend and into regular OPEX spend.
— Aaron Levie
QUESTIONS ANSWERED IN THIS EPISODE
5 questionsWhat specific tasks do you believe will remain irreducibly “human review” in regulated industries, even as models improve?
Levie argues the US–China AI dynamic is primarily a commercial and economic race, not an existential sprint where small timing advantages decide outcomes.
If junior legal and banking roles shrink, what concrete apprenticeship/mentorship structures replace the traditional pipeline?
He contends AI won’t eliminate most jobs because humans remain in the loop—AI shifts where review and accountability occur and reveals new bottlenecks that require more labor.
How should a Fortune 1000 company measure whether an ‘agent operator’ program is working—what KPIs matter beyond “time saved”?
A new “agent operator” role will emerge to redesign regulated enterprise workflows for agents, connect fragmented data systems, and manage ongoing model-driven workflow breakage.
Which software categories are most at risk of becoming ‘API-only commodities,’ and what moats (data, governance, logic) actually defend against that?
Software value will increasingly migrate from button-heavy UIs to robust APIs and embedded business logic, making “headless” platforms and strong governance layers essential.
What does an ‘agent-ready’ API look like in practice (auth, permissions, rate limits, provenance, audit logs, tool-calling design)?
Agent adoption will trigger a cybersecurity surge and a major budgeting shift as token/compute spend moves from IT budgets into broader operating expense tied to business outcomes.
Chapter Breakdown
AI “race” framing: commercial competition vs existential panic
Aaron reacts to the Jensen/Dwarkesh debate and argues the US–China AI narrative is often overstated. He frames AI as primarily a commercial and economic race (with safety considerations), not a binary, winner-take-all sprint where a few months decides everything.
Why “AI will take your job” is the wrong mental model
Aaron argues AI is shifting where humans enter workflows rather than removing humans entirely. He believes fear-driven narratives discourage people from entering fields society still needs, and he expects headcount growth in many professions even as tasks change.
More engineers (and even more lawyers): demand expands outside Big Tech
He predicts more engineers at Box and across the economy because most industries lack automation capacity. As non-tech sectors gain “Silicon Valley-grade” engineering leverage via AI coding tools, demand for technical talent broadens; similarly, legal content generation increases the need for qualified review and throughput.
The emerging role: “agent operator” and enterprise workflow redesign
Aaron outlines a new job category centered on implementing and maintaining agent-driven workflows inside regulated, complex enterprises. This role blends technical fluency with business process redesign and change management, acknowledging that enterprises can’t simply drop in agents without re-architecting processes and data.
Are SaaS products just databases in an agentic world? Where value moves
They explore whether agents will reduce SaaS to “valueless databases.” Aaron agrees some button-heavy software may commoditize, but argues durable value persists in APIs plus embedded business logic, governance, and human-in-the-loop collaboration experiences.
Agents as massive consumers/creators of unstructured data—and why platforms still matter
Aaron expects an explosion of unstructured content: contracts, marketing assets, reports, and analyses generated and reviewed by agents. He argues this increases the need for a governed backbone to store, secure, and manage that data—positioning systems like Box as force multipliers rather than commoditized repositories.
The cybersecurity tsunami: agents create (and must solve) new risk
Aaron describes a step-change in security risk as AI generates far more code than humans can realistically review. Agents also empower attackers to scan and exploit faster, creating a dynamic environment where agentic security becomes essential.
Token maxing and enterprise budgeting: from IT spend to line-of-business OpEx
Aaron explains how enterprises will allocate token/compute budgets and why ‘unlimited tokens’ is unrealistic for EPS-driven companies. He predicts token budgets shift from centralized IT budgets to business-unit operating budgets, enabling new ROI-based tradeoffs and potentially expanding total spend.
Enterprise adoption will be slower and services-heavy: change management and accountability
He argues diffusion will take longer than Silicon Valley assumes because enterprises face compliance, data fragmentation, and liability constraints. Professional services (Accenture-style and new specialists) will thrive by modernizing data estates and redesigning workflows for safe agent deployment.
Open-source and Chinese models in the stack: pragmatic usage with ongoing human oversight
They discuss the reality that companies benchmark frontier models and deploy cheaper open-source alternatives (often Chinese) to approach frontier performance. Aaron views this as empirically true but not inherently panic-inducing, while reiterating that even best models still need oversight.
Why many public-company agent products feel weak—and why this year is brutal execution
Aaron attributes weaker agent rollouts to the speed and complexity of the agent ecosystem: staying current requires deep practitioner-level engagement. He describes the CEO challenge of navigating rapid technical shifts while building a safe bridge for customers.
Monetizing agents in enterprise software: pricing, tiers, and Wall Street expectations
They debate whether agent products must re-accelerate revenue growth to matter publicly. Aaron says monetization is real, often via higher tiers that bundle automation/workflow value, but the market will wait to see which companies are structurally helped vs pressured by agents.
Frontier labs and the next infrastructure layer: why he’d still bet on the labs
Aaron says he would still invest heavily in frontier rounds, expecting valuations could go higher given market size. He also highlights emerging “picks-and-shovels” categories like agent observability and evals as broadly necessary across enterprises deploying agents.
EVERY SPOKEN WORD
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome