Loading video...

Video Failed to Load

Go Home

[NEW] a16z a16z speedrun 🧊 request for startups: Social Simulation We’ve trained AI to predict text, images, and even protein structures—but what if the next frontier is predicting us? What we’ll do, how we’ll act, what we’ll choose? Imagine launching products into synthetic markets before writing code. Testing political...

60,676 views • 1 year ago •via X (Twitter)

11 Comments

Colin Campbell's profile picture
Colin Campbell1 year ago

Note: This content is not a solicitation or offering for securities, nor should it be construed as investment advice. See for additional information. For a full list of portfolio companies, visit

ksminnovation's profile picture
ksminnovation1 year ago

AI is transforming healthcare! A KSM-led study shows AI can detect Celiac disease 4 years earlier @TalPatalon @MedPredict

Robin Guo's profile picture
Robin Guo1 year ago

@speedrun Accurate agentic simulations with non-determinism baked in will be super interesting

Colin Campbell's profile picture
Colin Campbell1 year ago

@speedrun Totally. A true crystal ball...

Doug McCracken's profile picture
Doug McCracken1 year ago

@speedrun can't wait to work with new teams building social simulations!

Colin Campbell's profile picture
Colin Campbell1 year ago

@speedrun Absolutely… so much to do in the space

Jordan Mazer's profile picture
Jordan Mazer1 year ago

@speedrun I think I've seen a movie about this

Colin Campbell's profile picture
Colin Campbell1 year ago

@speedrun What movie is this?

Troy Kirwin's profile picture
Troy Kirwin1 year ago

@speedrun The next pandemic will be simulated out with agents

Colin Campbell's profile picture
Colin Campbell1 year ago

@speedrun Hopefully the simulation just says "move to florida for 18 months" again

Samuel Spitz's profile picture
Samuel Spitz1 year ago

@speedrun Cc @khushkhushkhush they’re copy trading

Related Videos

In 2026, I believe we'll see a Consumer AI Renaissance. Consumer spending is the lifeblood of the US economy – representing more than 65% of US GDP. AI is already rapidly transforming how we spend both our money and time. In just the past 3 years, ChatGPT has become the fastest growing consumer product in history – reaching billions of users with engagement levels similar to social media apps like WhatsApp and Instagram. And with recent advancement in AI capabilities, I predict we’ll see a new wave of breakout consumer AI applications. We’ve seen major advances in three key areas: 1. Real-time voice AI - which allows users to engage with AI products in a natural, hands-free way 2. Advanced reasoning models - which improves the complexity of questions AI products can answer 3. Embedded Memory - which allows AI products to remember context between conversations, enabling more personalized responses With these new capabilities, we can now build AI products that are 10x better than before. I’m especially excited about products that use AI to make previously expensive services cheaper and more accessible, sometimes using human-in-the-loop to start. This includes things like AI travel agents, personal assistants, matchmakers, therapists, tutors, and more. We’ve already seen a number of early-stage teams working on these at a16z a16z speedrun 🧊 – and we’re excited to meet more. If you’re building the next generation of consumer AI, we’d love to hear from you! h/t to Theresa Horne Avenir whose excellent research helped inspire some of these insights.

Kenan Saleh

15,892 views • 8 months ago

Today, we’re excited to announce our $50M Series B, led by Greenfield Partners, with participation from Lightspeed and Notable Capital. 🚀 At Patronus AI, we develop simulations and evals to train and improve AI. The first phase of AI was built on static benchmarks, but that era is over. As agents are used to solve longer and longer tasks, they need to practice in dynamic, living worlds to get better. Simulations are the critical infrastructure powering this next phase. As a company, we’re behind the most influential research and products in AI evaluation, like FinanceBench, Lynx, and Percival. And things have moved at the speed of light since.⚡ We partner with the world's leading frontier AI labs and enterprises, and our revenue has grown more than 15x over the past year. Additionally, today, we’re introducing a preview of the first Digital World Model for AI agent training and simulation: Patronus-DWM. Digital World Models are language diffusion world models that predict realistic environment behaviors and steer agent actions across digital workflows. Just as physical world models predict how objects move through space, we’re developing the equivalent for the digital world: predicting how agents act in digital workflows, then using that to scale the creation of high-quality training data for LLMs. Digital World Models help us push the frontier of ultra long horizon workflows, and unlock a new class of self-improving RL environments. This is our scalable approach to simulating all of the world’s intelligence. The round was also joined by Datadog, Inc., Samsung Ventures, Gokul Rajaram, Factorial Capital, and a large cohort of amazing AI leaders across Anthropic, OpenAI, Google DeepMind, NVIDIA, Recursive, and more.✨ It has been the ride of a lifetime. But we’re just getting started. The best is yet to come. "Do not go gentle into that good night, Rage, rage against the dying of the light" - Dylan Thomas (1954)

PatronusAI

95,202 views • 2 months ago

Today, we’re excited to announce our $50M Series B, led by Greenfield Partners (formerly TPG Capital), with participation from Lightspeed and Notable Capital. 🚀 At PatronusAI, we develop simulations and evals to train and improve AI. The first phase of AI was built on static benchmarks, but that era is over now. As agents are used to solve longer and longer tasks, they need to practice in dynamic, living worlds to get better. Simulations are the critical infrastructure powering this next phase. As a company, we’re behind the most influential research and products in AI evaluation, like FinanceBench, Lynx, and Percival. And things have moved at the speed of light since. ⚡ We partner with the world's leading frontier AI labs and enterprises, and our revenue has grown more than 15x over the past year. Additionally, today, we’re introducing a preview of the first Digital World Model for AI agent training and simulation: Patronus-DWM. Digital World Models are language diffusion world models that predict realistic environment behaviors and steer agent actions across digital workflows. Just as physical world models predict how objects move through space, we’re developing the equivalent for the digital world: predicting how agents act in digital workflows, then using that to scale the creation of high-quality training data for LLMs. Digital World Models help us push the frontier of ultra long horizon workflows, and unlock a new class of self-improving RL environments. This is our scalable approach to simulating all of the world’s intelligence. The round was also joined by Datadog, Inc., Samsung Ventures, Gokul Rajaram, Factorial Capital, and a large cohort of amazing AI leaders and researchers across Anthropic, OpenAI, Google DeepMind, NVIDIA, Recursive, and more. ✨ It has been the ride of a lifetime. But we’re just getting started. The best is yet to come. "Do not go gentle into that good night, Rage, rage against the dying of the light" - Dylan Thomas (1954)

Anand Kannappan

40,406 views • 2 months ago

🚨 a16z a16z speedrun 🧊 applications are now open! we invest up to $1M in new start-ups building at the frontier - but that's just the start. beyond the investment, founders join a legendary community, work with our world-class operating team, and tap into the a16z network. details below: 1) the founder community - starting a new company is an emotional rollercoaster - being able to talk to founders on a similar journey can make it a bit less lonely - Speedrun is a tight-knit community of founders who support one another closely - providing advice, design partners, talent referrals, even investing in one another’s startups - during the program, we open up the a16z offices in SF and LA to founders as coworking space. dozens of founders work alongside each other before establishing their own offices - we have an alumni community >600 strong that come together every quarter through reunions and events at Tech Week 2) the a16z operating team - our operating team is a team of experts at the top of their field who provide specialized programs and hands-on support to founders. a few examples of how we help Speedrun founders: a) recruiting support to assemble and scale your team - direct intros to top talent in AI research, ML, full stack engineering, etc b) people support on US immigration and comp guidance for new hires / advisors c) GTM support on finding design partners and over $5M in credits on cloud, AI, and software spend d) marketing programs to build your brand and craft the perfect launch video to showcase your product e) capital networks support to refine your pitch deck and connect you to hundreds of prospective investors at demo day - many founders consider our operating team to be an extension of their own team in impact and how closely they work together 3) the a16z network - a16z as a firm has massive breadth and scale - we invest in and have networks across AI, bio/health, consumer, crypto, enterprise, fintech, games, infrastructure, american dynamism, and more - our network provides our founders an unfair advantage - it’s a rolodex they can tap into for intros to potential customers, employees, partners, and investors - providing a sense of power and reach you dont usually build until you're much farther along we know the best founders have many choices. if you’re setting out to build a generational company - we're confident you'll get unparalleled operating support, networks, & community from a16z Speedrun - all from day 0 if this sounds exciting, check out the link below!

Jon Lai

784,248 views • 11 months ago

🚨 FOMO Is Shaping the Future of AI – The Launch Is Almost Here! 🚨 Imagine a world where AI agents aren’t just bots—they’re fully autonomous, living personalities capable of learning, engaging, and creating across multiple platforms. FOMO’s new AI launchpad on Solana is here to make that future a reality. 🌐 Starting with our first Initial Agent Offering (IAO), FOMO is unleashing a new generation of AI agents that will redefine digital interaction: - On-Chain AI – Agents that are decentralized, fully autonomous, and ready to interact in real-time. - Multiplatform Presence – From X and Telegram to TikTok and YouTube, these agents are social media natives with a mission. - Real-Time Learning & Engagement – Agents will evolve and improve as they interact, shilling their tokens, creating content, and even performing complex tasks. FOMO’s Vision: This isn’t just AI; it’s the beginning of a movement that merges personality with purpose. By launching AI agents that can both engage and create, FOMO is opening doors to a world where digital personas can operate autonomously, driving value and utility in every interaction. Our pre-sale is still live for a limited time, but this is just the beginning of what FOMO is bringing to the space. Join us and become part of the AI agent revolution! 🔗 Join the Pre-Sale Now: We’re bringing you the future of crypto and AI—don’t blink, or you might miss the start of something legendary.

FOMO

21,882 views • 1 year ago

We’re entering a world of infinite output. AI can generate code, designs, and create even entire product roadmaps in seconds. Building is getting cheaper by the day. But judgment isn’t. It’s no longer about who can produce more. It’s about who can decide better. Taste and discernment isn’t just aesthetics. It’s: - Cutting a paragraph that sounds smart but says nothing - Knowing which slide to delete, not which one to add - Choosing the one priority that actually moves the needle And ignoring the flood of “good suggestions” that dilute focus. The hardest part of most white-collar work isn’t generating options anymore. It’s choosing well. And yet most tools optimize for speed and volume, not helping people get better at deciding. I’d love to see startups building toward products that help individuals train better judgement over time. What would it look like to create: - Tools that train people to attach real probabilities to their beliefs - and see, over time, where they tend to overestimate or underestimate - Systems that can make decision-making visible, trackable, and improvable - Tools that train you to prioritize ruthlessly, that is to pick the few things that matter - and get comfortable killing the rest In an age of infinite leverage, leverage without taste creates noise. I’m especially excited about founders building products that strengthen human discernment. In a world of abundance, discernment is the scarce asset. If you’re building here, DM me and let’s talk!

Lejla Johnsen

18,607 views • 3 months ago