

542 episodes
by Andreessen Horowitz · English
The a16z Show discusses tech and culture trends, news, and the future – especially as ‘software eats the world’. It features industry experts, business leaders, and other interesting thinkers and voices from around the world. This show is produced by Andreessen Horowitz (aka “a16z”), a Silicon Valley-based venture capital firm. Multiple episodes are released every week; visit a16z.com for more details and to sign up for our newsletters and other content as well!

A16Z’s AD50 episode on how startups are rebuilding America’s industrial base end-to-end (inputs, factories, and defense/robotics/commercial systems).
29 Oct 2026 · 50m · 23 chapters
Mark Cuban argues U.S. healthcare is “not efficient” because it lacks competitive dynamics and has broken incentives.
28 Oct 2026 · 1h 1m · 28 chapters
Reinventing K-12 and university education for the AI era by redesigning the school day around a “learning engine” (powered by cognitive/learning science and LLMs) plus real-world projects that build…
27 Oct 2026 · 57m · 23 chapters
OpenAI’s computer-using AI agents (Astra, Sol, and “Dots”) that can operate software faster than humans, with enterprise controls for safety, auditability, and human approval loops; future focus on…
26 Oct 2026 · 45m · 17 chapters
How AI agents change startup leverage, strategy, and competitive dynamics; why distribution still matters; and where new agent startups will emerge across industries.
25 Oct 2026 · 34m · 10 chapters
Ben Horowitz (A16Z co-founder) explains what it takes to win as a founder in the AI era—what AI changes (faster scaling to massive audiences, easier software creation) and what it doesn’t (founder…
24 Oct 2026 · 42m · 13 chapters
Amazon CEO Andy Jassy explains why Amazon expects to spend about $220B in 2026 CapEx to meet AI infrastructure demand, arguing demand is still outstripping supply and that AWS could become a $1T…
23 Oct 2026 · 1h 1m · 23 chapters
a16z expands its crypto fund into “A16Z Frontier,” arguing AI is widening the gap between what early adopters can build and what most companies can, so founders should be “ambitious enough” (not “too…
22 Oct 2026 · 1h 1m · 30 chapters
The infrastructure required for AI agents to reliably work with documents (PDFs, spreadsheets, scans), arguing that “accurate” models can still silently drop information and that OCR alone is…
21 Oct 2026 · 38m · 18 chapters
OpenAI health AI research vision for making trustworthy, low-cost “health intelligence” widely accessible, and how AI could act as a safety net/quarterback for patients and clinicians while reducing…
20 Oct 2026 · 56m · 29 chapters
Town (Town.com) is an AI assistant for work that automates knowledge-worker workflows (email, calendar, meeting prep, drafting responses) and can act on users’ behalf.
19 Oct 2026 · 1h 3m · 21 chapters
Moving AI from benchmarks to production for high-stakes industries (housing and healthcare), where small error rates become unacceptable at massive scale.
18 Oct 2026 · 24m · 12 chapters
Healthcare scheduling access is still slow (weeks to book), and the episode argues AI can fix it by making appointment booking “actionable” across many “front doors” (websites, payers, and AI agents).
17 Oct 2026 · 35m · 17 chapters
Whether AI that can generate production code can also run and troubleshoot real operational systems (distributed clusters, observability/telemetry, AIOps/AISRE, and security).
16 Oct 2026 · 1h · 23 chapters
How “agentic commerce” changes payments when AI agents can spend via payment credentials, and how credential vaulting/tokenization (VGS) enables it while preserving PCI/chargeback protections.
15 Oct 2026 · 51m · 25 chapters
Race between open and closed AI “frontier” models; whether open-weight models can keep up, and what drives progress (compute, data, talent).
14 Oct 2026 · 1h · 40 chapters
Whether recent AI math breakthroughs (e.g., OpenAI results) could lead to faster breakthroughs against deployed cryptography, potentially breaking assumptions behind systems like elliptic-curve…
13 Oct 2026 · 42m · 27 chapters
Recursive self-improvement (RSI) in AI—distinguishing true closed-loop RSI from “AI-assisted automation,” and arguing that real progress must be measured by both capability gains (numerator) and…
12 Oct 2026 · 33m · 14 chapters
Nick Bostrom revisits AI alignment and existential risk in light of today’s LLMs, arguing for “pacing the frontier” (keep capability development moving but preserve the option to slow down) rather…
11 Oct 2026 · 54m · 19 chapters
Ben Horowitz (Andreessen Horowitz co-founder) explains what makes a great entrepreneur and how investors should evaluate founders: look for world-class strengths, not the absence of weaknesses.
10 Oct 2026 · 52m · 20 chapters
AI infrastructure bottlenecks beyond chip design—energy/power limits, memory bandwidth, manufacturing timelines, and networking/optics for large GPU clusters.
9 Oct 2026 · 54m · 16 chapters
AWS infrastructure for agentic AI workflows—how AWS is optimizing cloud services, security, and performance for agents; how it plans GPU capacity amid massive demand; and how custom silicon…
8 Oct 2026 · 56m · 21 chapters
Valen’s effort to rebuild mortgage servicing infrastructure from scratch—replacing legacy, pre-internet systems in a $13T consumer-debt market—then using that “system of record” plus AI agents to…
7 Oct 2026 · 41m · 15 chapters
How AI changes cyber offense/defense, why defense must become autonomous, and how Armiden builds “AI on offense” to continuously find exploitable risk and “AI on defense” to apply compensating…
6 Oct 2026 · 49m · 20 chapters