a16zReplit CEO Amjad Masad on What Young People Should Learn in the AI Era
At a glance
WHAT IT’S REALLY ABOUT
AI-era education should cultivate heretical curiosity, projects, and character
- Amjad Masad argues the core societal value of young people is to be “heretics” who question entrenched assumptions, while modern college often trains conformity and risk aversion.
- The conversation sketches an ideal 18–22 pathway centered on freedom to explore, supportive resources for “side quests,” and peer groups that reinforce unconventional thinking rather than credential compliance.
- They advocate project-based learning as the default model—made newly scalable by AI—so students learn skills through building things they care about instead of following abstract syllabi.
- Masad and hosts warn that early startup fundraising can be “premature optimization,” driven by envy and predatory capital, and propose curiosity-led exploration as a healthier precursor to commitment.
- Beyond skills, they emphasize discernment, interpersonal competence, trust-based culture, and virtue-oriented ethics as essential preparation for leadership in rapidly changing AI-era institutions.
IDEAS WORTH REMEMBERING
5 ideasYoung people’s “job” is heresy: question deeply held beliefs and explore the unknown.
Masad uses a sci‑fi allegory (“The City and the Stars”) to argue that societies stagnate when they compress novelty and train people to comply; young people’s unique advantage is fresh perspective and willingness to challenge taboos. He claims many college environments do the reverse by socializing students into conformity and credential-chasing.
Design education around freedom, meandering exploration, and high-quality peers—not rigid requirements.
They argue an ideal 18–22 experience would maximize optionality: explore unfamiliar subjects (Jobs’ calligraphy example), follow “side quests,” and receive institutional support (labs, compute, mentors) rather than penalties (attendance mandates, rigid tracks). The peer environment matters: surround yourself with non‑conventional thinkers or you’ll become conventional.
Project-based learning + AI unlocks practical, customized education at scale.
They critique learning detached from use (“why am I learning this?”) and advocate project-based learning as the default: build something you care about and learn the tool (Python, stats, etc.) in context. AI makes personalized, self-directed projects more feasible because students can scaffold their own curriculum with tools like ChatGPT/Claude.
Starting a startup too early can be premature optimization; curiosity should lead, not envy or hype.
Masad frames “raising money at 18” as often premature optimization: kids get pulled into startups by status/envious comparison, pre-commit to ideas they don’t love, and accept predatory pre-idea capital that constrains them early. The healthier signal is an idea with unavoidable gravity—momentum that “forces” commitment because it’s intrinsically compelling.
The meta-skill is discernment: when to follow rules and when to break them (and how to work with people regardless).
They discuss a balance between disruption and social functioning: be contrarian enough to generate new ideas, but skilled enough interpersonally to collaborate. Masad emphasizes discernment—knowing when to trust elders/systems and when to override them—built through “reps,” autonomy, and learning from failure (asymmetric upside vs downside).
WORDS WORTH SAVING
5 quotesI think ultimately the utility of young people is to question deeply held beliefs, is to be heretics.
— Amjad Masad
And in today's college, it's actually the opposite. It kind of beats any outside thought out of them.
— Amjad Masad
This, like, meandering that I think s- you know, modern society, uh, thinks is sort of useless, I think is actually the most useful in college.
— Amjad Masad
Once you find that idea that you wanna dedicate your life to or dedicate the next 10 years to, it's gonna have such gravity you're not gonna be able to avoid it.
— Amjad Masad
I thi- I think the first thing is that they dove really deep in something. They can teach me something.
— Amjad Masad
High quality AI-generated summary created from speaker-labeled transcript.
