Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Ben Horowitz on foundation models as infrastructure: “People believed the big foundation models would be giant brains that could do anything better than anybody.” “It has not played out quite like that.” “Maybe the application behavior is actually more important than having the biggest model trained with the most... show more

a16z

937,114 subscribers

46,594 views • 7 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Ben Horowitz on AI: “This is a bigger technology market than I've ever seen.” In this exclusive conversation with a16z's Jen Kha, Ben covers: - Why he doesn't think there's an AI bubble - How a16z is structured - The origin of our American Dynamism vertical - His most listened song in 2025 0:00 Introduction 1:33 Managing GPs vs companies 4:33 How a16z is structured 10:14 Culture and staying in the details 12:46 How to identify the right markets 21:28 M&A landscape opening up 22:09 Why foundation models alone aren't enough 25:46 Ownership and the future of VC 29:03 Why AI will produce more winners than previous technology cycles 32:01 Rapid-fire personal questions benahorowitz.eth Jen Kha

Ben Horowitz on AI: “This is a bigger technology market than I've ever seen.” In this exclusive conversation with a16z's Jen Kha, Ben covers: - Why he doesn't think there's an AI bubble - How a16z is structured - The origin of our American Dynamism vertical - His most listened song in 2025 0:00 Introduction 1:33 Managing GPs vs companies 4:33 How a16z is structured 10:14 Culture and staying in the details 12:46 How to identify the right markets 21:28 M&A landscape opening up 22:09 Why foundation models alone aren't enough 25:46 Ownership and the future of VC 29:03 Why AI will produce more winners than previous technology cycles 32:01 Rapid-fire personal questions benahorowitz.eth Jen Kha

a16z

99,934 views • 7 months ago

📁 Ben Horowitz says large foundation models are not universal brains. They are infrastructure. Real value comes from combining many models for specific use cases. Cursor proves it by running over 13 models and building its own coding model. The edge is no longer size, but orchestration.

📁 Ben Horowitz says large foundation models are not universal brains. They are infrastructure. Real value comes from combining many models for specific use cases. Cursor proves it by running over 13 models and building its own coding model. The edge is no longer size, but orchestration.

Jon Hernandez

14,503 views • 7 months ago

Behavioral Foundation Models (BFMs) trained with RL are secretly more powerful than we think. BFM’s directly output a policy believed to be near-optimal given any reward function. Our new work shows that they can actually do much better:

Behavioral Foundation Models (BFMs) trained with RL are secretly more powerful than we think. BFM’s directly output a policy believed to be near-optimal given any reward function. Our new work shows that they can actually do much better:

Harshit Sikchi

44,182 views • 1 year ago

Ben Horowitz once told Martin Casado that pricing is the most important decision in the history of a company. Pricing determines both growth and margins, the two forces that drive valuation. Martin says the same lesson still applies today as software moves from recurring licenses to usage-based models. Source: martin_casado with Mario Gabriele 🦊 on The Generalist

Ben Horowitz once told Martin Casado that pricing is the most important decision in the history of a company. Pricing determines both growth and margins, the two forces that drive valuation. Martin says the same lesson still applies today as software moves from recurring licenses to usage-based models. Source: martin_casado with Mario Gabriele 🦊 on The Generalist

a16z

44,691 views • 10 months ago

"As roboticists and computer vision people [outside of big tech], do we just wait for the next foundation model?" I share the frustration. It's disempowering when most major progress recently is downstream of "foundation models" that you don't have the compute or data to train.

"As roboticists and computer vision people [outside of big tech], do we just wait for the next foundation model?" I share the frustration. It's disempowering when most major progress recently is downstream of "foundation models" that you don't have the compute or data to train.

Chris Offner

68,742 views • 1 year ago

Being a leader is saying what people don’t want to hear. Ben Horowitz: “If everyone agrees with the decision, then you didn’t add any value because they would’ve done that without you.” “The only value you ever add is when you make a decision that most people don’t like.” benahorowitz.eth Lenny Rachitsky

Being a leader is saying what people don’t want to hear. Ben Horowitz: “If everyone agrees with the decision, then you didn’t add any value because they would’ve done that without you.” “The only value you ever add is when you make a decision that most people don’t like.” benahorowitz.eth Lenny Rachitsky

a16z

47,102 views • 11 months ago

Balaji Srinivasan proposes a western version of the special economic zone where builders can operate at the speed of physics, not permits. “The special Elon zone, or more generally, the special founder zone... It’s like an essentially uninhibited territory, as close to Burning Man as possible” Ben Horowitz: “I think that’s very doable. There’s so many states in the US that would wanna do that. I know Nevada would wanna do that.” Balaji and benahorowitz.eth on Network State podcast

Balaji Srinivasan proposes a western version of the special economic zone where builders can operate at the speed of physics, not permits. “The special Elon zone, or more generally, the special founder zone... It’s like an essentially uninhibited territory, as close to Burning Man as possible” Ben Horowitz: “I think that’s very doable. There’s so many states in the US that would wanna do that. I know Nevada would wanna do that.” Balaji and benahorowitz.eth on Network State podcast

a16z

301,291 views • 6 months ago

Marc Andreessen and Ben Horowitz explain the problem with the conventional wisdom of judging markets on how they look today: “As [we] go through our careers, we just see more and more examples where the mistakes that we or others make is, - The market for Uber and Lyft must be the market for taxi cabs, - Or the market for cloud software must be the same as the market for on-prem software, - Or the market for GPUs must be the market for people who like to play games.” The winners build technology that unlocks something on the supply side and grows market size by orders of magnitude. Marc Andreessen 🇺🇸 benahorowitz.eth

Marc Andreessen and Ben Horowitz explain the problem with the conventional wisdom of judging markets on how they look today: “As [we] go through our careers, we just see more and more examples where the mistakes that we or others make is, - The market for Uber and Lyft must be the market for taxi cabs, - Or the market for cloud software must be the same as the market for on-prem software, - Or the market for GPUs must be the market for people who like to play games.” The winners build technology that unlocks something on the supply side and grows market size by orders of magnitude. Marc Andreessen 🇺🇸 benahorowitz.eth

a16z

68,941 views • 7 months ago

"AI is an operating system." Lux Capital's Deena Shakir says the biggest opportunities are shifting to application layers like healthcare and robotics, where startups can build on top of foundation models rather than compete with them

"AI is an operating system." Lux Capital's Deena Shakir says the biggest opportunities are shifting to application layers like healthcare and robotics, where startups can build on top of foundation models rather than compete with them

Bloomberg TV

18,169 views • 22 days ago

“We are dream builders. We’re not dream killers.” Ben Horowitz on a16z's written culture that forbids publicly attacking founders or startups: “We have a culture that is written culture... You cannot join the firm unless you sign the culture document that says you'll adhere to this culture.” “Our whole point is, if you want to do something larger than yourself and make the world a better place, we are a hundred percent for it.” benahorowitz.eth Packy McCormick

“We are dream builders. We’re not dream killers.” Ben Horowitz on a16z's written culture that forbids publicly attacking founders or startups: “We have a culture that is written culture... You cannot join the firm unless you sign the culture document that says you'll adhere to this culture.” “Our whole point is, if you want to do something larger than yourself and make the world a better place, we are a hundred percent for it.” benahorowitz.eth Packy McCormick

a16z

395,938 views • 7 months ago

Exa CEO Will Bryk explains why retrieval can help solve the tokenpocalypse: "We should not be using gigantic models for every task." "You should use a family of models of different sizes. The big model decides what to do, and it dishes out commands to the small models, and those small models can be way more accurate and reliable if they're using retrieval." "Retrieval helps small models act like big models... We do save our customers a huge amount of tokens because they can use smaller models and use retrieval." "We could save 20x on cost for customers compared to other providers by being very efficient in what information from the web the agent actually sees." Will Bryk with Sarah Wang

Exa CEO Will Bryk explains why retrieval can help solve the tokenpocalypse: "We should not be using gigantic models for every task." "You should use a family of models of different sizes. The big model decides what to do, and it dishes out commands to the small models, and those small models can be way more accurate and reliable if they're using retrieval." "Retrieval helps small models act like big models... We do save our customers a huge amount of tokens because they can use smaller models and use retrieval." "We could save 20x on cost for customers compared to other providers by being very efficient in what information from the web the agent actually sees." Will Bryk with Sarah Wang

a16z

25,144 views • 2 months ago

Goldman Sachs CEO David Solomon on the open question of agentic investing: "Models work on the facts that are available." "Sometimes the biggest changes in the way you have to think about investing in a portfolio comes from things that are completely new and unexpected." "It can't be something from the past." "There are a handful of people who have so outperformed as investors over a long period of time, but... generally speaking, people underperform." "And so if the models are based on the information that the people are underperforming all have, it's going to be interesting to see whether something different comes out of it." David Solomon with David Haber and benahorowitz.eth

Goldman Sachs CEO David Solomon on the open question of agentic investing: "Models work on the facts that are available." "Sometimes the biggest changes in the way you have to think about investing in a portfolio comes from things that are completely new and unexpected." "It can't be something from the past." "There are a handful of people who have so outperformed as investors over a long period of time, but... generally speaking, people underperform." "And so if the models are based on the information that the people are underperforming all have, it's going to be interesting to see whether something different comes out of it." David Solomon with David Haber and benahorowitz.eth

a16z

62,860 views • 6 months ago

Ben Horowitz: "Was Steve Jobs so mean to people because that just goes with being so brilliant? Or because he could get away with it, because he was so brilliant?" "I think it's a little bit more the latter. He didn't have to do that." "The problem with founders modeling themselves after Jobs... Well, if you're not Jobs or Elon, maybe you can't get away with that." "It's very hard to have the highest performing talent without them coming with some of those behaviors." benahorowitz.eth with Brian Halligan

Ben Horowitz: "Was Steve Jobs so mean to people because that just goes with being so brilliant? Or because he could get away with it, because he was so brilliant?" "I think it's a little bit more the latter. He didn't have to do that." "The problem with founders modeling themselves after Jobs... Well, if you're not Jobs or Elon, maybe you can't get away with that." "It's very hard to have the highest performing talent without them coming with some of those behaviors." benahorowitz.eth with Brian Halligan

a16z

72,841 views • 5 months ago

Ben Horowitz: "It's the best thing in the world for people to try to do something larger than yourself." "Why wouldn't you want to fund as much of that as you can?" Read The Case for Scaling Venture:

Ben Horowitz: "It's the best thing in the world for people to try to do something larger than yourself." "Why wouldn't you want to fund as much of that as you can?" Read The Case for Scaling Venture:

a16z

36,393 views • 6 months ago

Gemma 4 is a bigger deal than more people realize... It's an incredible model that fits on most consumer hardware. The future is hybrid hosted/edge models.

Gemma 4 is a bigger deal than more people realize... It's an incredible model that fits on most consumer hardware. The future is hybrid hosted/edge models.

Matthew Berman

52,946 views • 4 months ago

Ben Horowitz on the importance of culture: "A culture is not a set of ideas, it’s a set of actions." "If you define your culture as a kind of set of ideas—integrity, do the right thing, we have each other’s backs, or any of these ‘corporate values’—it’s actually just a bunch of platitudes. It doesn’t mean anything. The culture has to be defined in terms of the exact behavior you want that support that idea." "It’s the little things—how responsive are you to your colleagues? What’s the SLA on returning a Slack message or an email? Do you show up to meetings on time?" "We have an idea about—you have to respect the entrepreneur... you can’t ever be late to a meeting with an entrepreneur. I used to fine people $10/minute in the beginning of the firm to reinforce it." Source: benahorowitz.eth on Invest Like the Best with Patrick OShaughnessy

Ben Horowitz on the importance of culture: "A culture is not a set of ideas, it’s a set of actions." "If you define your culture as a kind of set of ideas—integrity, do the right thing, we have each other’s backs, or any of these ‘corporate values’—it’s actually just a bunch of platitudes. It doesn’t mean anything. The culture has to be defined in terms of the exact behavior you want that support that idea." "It’s the little things—how responsive are you to your colleagues? What’s the SLA on returning a Slack message or an email? Do you show up to meetings on time?" "We have an idea about—you have to respect the entrepreneur... you can’t ever be late to a meeting with an entrepreneur. I used to fine people $10/minute in the beginning of the firm to reinforce it." Source: benahorowitz.eth on Invest Like the Best with Patrick OShaughnessy

a16z

178,793 views • 6 months ago

Ben Horowitz and Alex Rampell on AI Infrastructure, Crypto, and the Future of Venture a16z's Ben Horowitz and Alex Rampell sit down to discuss why the fundamental rules of building software companies have changed, how AI is creating massive demand for crypto infrastructure, why the next 15 years will make everyone's lives dramatically better, and more. 00:00 Intro 01:39 The new laws of physics for tech companies 06:37 Why not every legacy SaaS company is dead 10:30 America's AI infrastructure bottleneck 14:27 AI + crypto: Why they're more connected than you think 19:54 The future of VC 24:06 AI and the history of technological change benahorowitz.eth Alex Rampell

Ben Horowitz and Alex Rampell on AI Infrastructure, Crypto, and the Future of Venture a16z's Ben Horowitz and Alex Rampell sit down to discuss why the fundamental rules of building software companies have changed, how AI is creating massive demand for crypto infrastructure, why the next 15 years will make everyone's lives dramatically better, and more. 00:00 Intro 01:39 The new laws of physics for tech companies 06:37 Why not every legacy SaaS company is dead 10:30 America's AI infrastructure bottleneck 14:27 AI + crypto: Why they're more connected than you think 19:54 The future of VC 24:06 AI and the history of technological change benahorowitz.eth Alex Rampell

a16z

24,904 views • 4 months ago

Ben Horowitz on the standout skill required to be an exceptional CEO: Don't hesitate to make the hard decision when you see the truth. “You could be really, really smart, but if you wait too long before you pull the trigger, you’re not smart anymore.” The real killer is the non-decision. As Ben once said: “You need to run towards the pain and darkness and not away from it." Source: benahorowitz.eth on My First Million with Shaan Puri and Sam Parr

Ben Horowitz on the standout skill required to be an exceptional CEO: Don't hesitate to make the hard decision when you see the truth. “You could be really, really smart, but if you wait too long before you pull the trigger, you’re not smart anymore.” The real killer is the non-decision. As Ben once said: “You need to run towards the pain and darkness and not away from it." Source: benahorowitz.eth on My First Million with Shaan Puri and Sam Parr

a16z

96,124 views • 8 months ago

The nervous system can be trained. And there is no better time to start training it than yesterday. Love this time with that dude more than anything. Hammers Only.

The nervous system can be trained. And there is no better time to start training it than yesterday. Love this time with that dude more than anything. Hammers Only.

Ray Zingler

25,777 views • 9 months ago

“Maybe we are at the last model that humans could have figured out.” Ex-OpenAI researcher Jerry Tworek says AI agents may soon design better models than humans. The search space may be too large for us alone.

“Maybe we are at the last model that humans could have figured out.” Ex-OpenAI researcher Jerry Tworek says AI agents may soon design better models than humans. The search space may be too large for us alone.

vitrupo

77,272 views • 6 months ago