Skip to content
a16za16z

How AI Agents Will Transform in 2026 (a16z Big Ideas)

AI is moving from chat to action. In this episode of Big Ideas 2026, we unpack three shifts shaping what comes next for AI products. The change is not just smarter models, but software itself taking on a new form. You will hear from Marc Andrusko on the shift from prompting to execution, Stephanie Zhang on what it means to build machine-legible software, and Olivia Moore on why voice agents are becoming practical, deployable systems rather than demos. Together, these ideas tell a single story. Interfaces shift from chat to action, design shifts from human-first to agent-readable, and work shifts to agentic execution. AI stops being something you ask, and becomes something that does. Timecodes: 0:00 Introduction: The Future of AI Interfaces 0:30 The Death of the Prompt Box 1:09 AI as the Ultimate Employee 2:28 Proactive AI in CRM and Workflows 4:09 Designing for Agents, Not Humans 5:28 Machine Legibility and Content Creation 8:48 The Rise of AI Voice Agents 9:25 Voice AI in Healthcare, Finance, and Recruiting 11:01 Challenges and Opportunities in Voice AI 12:32 Consumer Voice AI and Wellness 13:01 Building with Voice AI: Tools and Platforms Resources: Follow Marc Andrusko on X: https://twitter.com/mandrusko1 Follow Stephanie Zhang on X: https://twitter.com/steph_zhang Follow Olivia Moore on X: https://twitter.com/omooretweets Read more of our 2026 Big Ideas Part 1: https://a16z.com/newsletter/big-ideas-2026-part-1 Part 2: https://a16z.com/newsletter/big-ideas-2026-part-2 Part 3: https://a16z.com/newsletter/big-ideas-2026-part-3 Stay Updated: If you enjoyed this episode, be sure to like, subscribe, and share with your friends! Find a16z on X: https://x.com/a16z Find a16z on LinkedIn: https://www.linkedin.com/company/a16z Listen to the a16z Podcast on Spotify: https://open.spotify.com/show/5bC65RDvs3oxnLyqqvkUYX Listen to the a16z Podcast on Apple Podcasts: https://podcasts.apple.com/us/podcast/a16z-podcast/id842818711 Follow our host: https://x.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details, please see a16z.com/disclosures.

Erik TorenberghostMarc AndruskoguestStephanie ZhangguestOlivia Mooreguest
Dec 21, 202513mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

AI apps shift from prompting to proactive agent-led work experiences

  1. AI apps are moving beyond the prompt box toward proactive systems that observe work, propose actions, and require only final human approval in many cases.
  2. The economic upside shifts from capturing software spend to capturing labor spend, reframing AI apps as “AI employees” that can execute end-to-end tasks.
  3. Product creation and content strategy will increasingly optimize for agent consumption—machine legibility, relevance, and structured information—rather than human attention and visual UI.
  4. Enterprises are already deploying AI voice agents at scale in healthcare, finance, and recruiting due to staffing constraints, compliance consistency, and improving latency/accuracy.
  5. A key tension across agentic systems is where humans stay in the loop versus where autonomy is acceptable, especially in high-liability or high-stakes workflows.

IDEAS WORTH REMEMBERING

5 ideas

The next AI interface is proactive, not conversational.

Instead of asking users to craft prompts, agentic apps will monitor context (email, calendar, pipeline, telemetry) and surface recommended actions that users approve—especially in business workflows.

AI expands the market from software budgets to labor budgets.

The speakers frame AI’s opportunity as shifting from hundreds of billions in software spend to trillions in labor spend, because agents aim to do the actual work, not just manage information.

“High-agency employee” behavior is the blueprint for agents.

The target agent identifies problems, researches causes, evaluates solutions, implements a fix, and only requests approval at the end—mirroring top-tier human operators rather than junior task-takers.

CRMs will become continuously running deal-execution engines.

An AI-native CRM won’t just store pipeline data; it will mine historical emails/leads, draft re-engagement, prioritize next actions, and keep the funnel moving with minimal manual navigation.

Design and content must become machine-legible first.

As agents intermediate the web and apps, optimization shifts away from visual hierarchy and hooks toward clear structure, relevance, and extractable facts that agents can reliably parse and summarize.

WORDS WORTH SAVING

5 quotes

My big idea for twenty twenty-six is the death of the prompt box as the primary user interface for AI applications.

Marc Andrusko

The next wave of apps will require way less prompting. They'll observe what you're doing and intervene proactively with actions for you to review.

Marc Andrusko

The opportunity we're attacking used to be the three hundred to four hundred billion dollars of software spend annually in the world. Now what we're excited about is the thirteen trillion dollars of labor spend that exists in the US alone.

Marc Andrusko

We're no longer designing for humans, but for agents. The new optimization isn't visual hierarchy, but machine legibility.

Stephanie Zhang

You would think there's so much compliance and regulation that voice AI can't operate there yet, but it turns out this is an area where voice AI actually outperforms because humans are actually very good at violating compliance and regulations, and voice AI can get it every time, and importantly, you can track how voice AI is performing over time.

Olivia Moore

Death of the prompt boxAI as an “employee” and labor TAM expansionProactive workflows (CRM, email, calendar, call notes)Human-in-the-loop vs full autonomyDesigning for agents vs humansMachine legibility and GEO (LLM discovery/optimization)Voice agent platforms and vertical adoption

High quality AI-generated summary created from speaker-labeled 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