Marc Andreessen's 2026 Outlook: AI Timelines, US vs. China, and The Price of AI

What makes a startup worth acquiring when software has never been easier to build? Salesforce’s John Somorjai says the answer goes well beyond technology: the best acquisitions can accelerate a product roadmap, bring in exceptional talent, and put a startup in front of 15,000 Salesforce sales reps almost overnight. John joins a16z General Partner Alex Immerman to unpack the thinking behind nearly $100 billion in acquisitions during his tenure, from Slack and MuleSoft to a new generation of AI-native companies. They discuss Salesforce’s build-versus-buy framework, why outcome-based pricing is replacing seats in parts of enterprise software, and what founders should consider when deciding whether to sell. John also shares what can kill a deal during diligence, why he thinks parts of today’s venture market are overvalued, and how Salesforce Ventures is expanding from enterprise software into frontier AI and robotics.

AI is moving finance from reconciling what happened last month to helping decide what the business should do next. Figma CFO Praveer Melwani argues that the best finance teams are becoming builders themselves, automating repetitive work so they can spend more time understanding customers, running experiments, and shaping strategy. Praveer joins a16z’s Seema Amble and Ivan Makarov to share what that shift looks like inside Figma, from collections and expense analysis to M&A diligence, where the corp dev team now uses AI to prototype products and simulate how Figma leaders might challenge an investment memo. They discuss why AI “slop” is a real risk, how Figma decides what to build versus buy, and why successful internal experiments eventually need a centralized systems layer to turn one-off agents into reliable infrastructure. They also get into what this means for hiring the next generation of finance leaders. Praveer’s bar isn’t simply whether someone can use AI. It’s whether they can understand the business, interrogate the output, communicate clearly, and use automation without outsourcing their thinking.

Will we have one AI assistant that does everything, or an entire network of agents that know different parts of our lives? David Pawlan, creator of Assistant Benchmark, returns with Anish Acharya and Josh Elman to take stock of how quickly the agent landscape has changed. David has now tested more than 140 AI assistants, and they unpack what separates them, from memory and judgment to trust, shopping, and social experiences. They also explore what happens when agents start talking to each other, how they could create new kinds of social networks and serendipity, and why the most compelling agents may ultimately do more than save time. They could become a kind of secret superpower.

If you could do only three things for your health, what should they be? Andrew Huberman’s answer is surprisingly straightforward: get bright light shortly after waking, resistance train, and make the sleep you can get as high-quality as possible. Andrew joins a16z General Partner Daisy Wolf to walk through the protocols he actually follows, from exercise and nutrition to meditation, supplements, and locking his social media phone in a box. He also explains why optimizing every health protocol can be counterproductive if it gets in the way of the rest of your life. And for founders, Andrew makes a different case: identify the hours when your brain does its best work and protect them. They discuss cognitive performance, meditation, sustainable workloads, and emerging neurotechnology that could help us control transitions between sleep, focus, and recovery.

What does it take to build robots that can actually work in factories and warehouses? Tutor Intelligence CEO Josh Gruenstein argues robotics is a systems problem spanning AI, hardware, manufacturing, and the work itself. Josh joins a16z General Partners Martin Casado and Raghu Raghuram to explain Tutor’s approach: deploy robots into real businesses, use human “tutors” when needed, and turn that work into training data that makes the robots more capable. They discuss Tutor’s 100-robot data factory, robots starting at roughly $16 an hour, advances in spatial reasoning, and what flexible robotic labor could mean for American manufacturing and the physical economy.