View cryptocurrency as a foundational technology providing the essential economic and security network for an AI-driven world. The most promising long-term investments are blockchain projects focused on decentralized identity, data provenance to combat deepfakes, and secure computing using zero-knowledge proofs. Bitcoin (BTC) is considered a key foundational asset, acting as internet-native money for future AI agents. For another long-term opportunity, consider companies building out the domestic robotics supply chain, which is a critical strategic need for the US. Remember that both AI and robotics are multi-decade themes requiring a patient, long-term investment horizon.

Artificial Intelligence (AI)

• Ben Horowitz believes we are still very early in the AI technology cycle, which he sees as a 25-year arc. The next major breakthrough after the initial ones is still an "open question."
• The impact of AI is seen as both overrated and underrated. It will eventually affect every single sector, but in ways that are difficult to predict, similar to how the spreadsheet created the private equity industry.
• The products work exceptionally well, leading to unprecedented growth. OpenAI reached $10 billion in revenue in four years, which helps justify the massive valuations in the space.
• Valuations are extremely high in the private markets, with companies like Anthropic being valued at $183 billion and OpenAI potentially at half a trillion dollars.

Takeaways

• AI is a long-term, multi-decade investment theme, not a short-term trade. Investors should have a long time horizon.
• While the growth is real, the high valuations in leading private companies suggest that much of the future growth is already priced in. Public market investors should be cautious about valuations when these companies eventually IPO.
• The most significant opportunities may not be in the AI model creators themselves, but in the companies across various sectors (like entertainment, finance, defense, etc.) that successfully integrate AI to improve their economics and create new products.

Open Source AI

• A major theme is the competition between open source and closed source AI models.
• Horowitz argues that the US's strength is its open society, and the best way to compete with China is by embracing open source AI.
• He believes US policy has been "anti-open source," which he calls a "dumb mistake" that has allowed China to take the lead in this area.
• The dominant open source models, like DeepSeek, are now from China and are widely used in US academia and even by US companies.
• The "weights" of these models (the trained part of the AI) encode cultural values, and whoever has the dominant open source model can have a massive impact on global culture and information.

Takeaways

• The battle for AI dominance may be decided in the open-source arena, not just among proprietary giants like OpenAI and Google.
• Investors should pay close attention to the development of a strong US-based open-source AI ecosystem. Companies that successfully build and commercialize open-source models could be significant long-term winners.
• The geopolitical tension between the US and China is a critical factor. The success of US AI depends on policy shifts that favor open innovation rather than secrecy.

NVIDIA (NVDA)

• Mentioned in the context of US national security policy.
• The US has tried to stop the export of NVIDIA chips to China to slow their AI development.
• Horowitz believes this strategy "really hasn't" worked and that China's AI progress continues despite these restrictions.
• He suggests that such policies may end up retarding the growth of the US tech industry more than they harm competitors.

Takeaways

• The investment thesis for companies like NVIDIA often includes the idea that the US government will protect their technological lead through policy.
• The transcript suggests this may be a flawed assumption. The effectiveness of chip export bans is questionable, posing a potential risk to the long-term competitive moat of US chipmakers against a determined state actor like China.

Robotics

• Horowitz believes robotics will likely be the biggest industry in the future and is "super important."
• However, he thinks the timeline for advanced humanoid robots is "further away than anybody is saying."
• He uses the example of Waymo's self-driving cars, which took roughly 19 years to go from a proof-of-concept to a limited commercial service. Humanoid robotics is a much harder problem.
• Major Risk Factor: The entire robot supply chain is currently in China. This poses a significant strategic risk for the US, similar to its dependency on other countries for rare earth minerals and chips.

Takeaways

• Robotics is a very long-term investment. Investors should be patient and not expect a "humanoid robot in every home" in the next few years.
• A key investment opportunity could be in companies that work to build out a robotics supply chain in the US. This is a critical national security and economic need, and companies that solve this problem could receive significant government and private support.

Cryptocurrency & Blockchain

• Presented as a critical and complementary technology to AI, forming the "network" pillar to AI's "computer" pillar.
• Bitcoin (BTC) is mentioned as a foundational technology, with its whitepaper being described as Nobel Prize-worthy.
• Bullish Sentiment: Horowitz sees crypto and blockchain as providing essential infrastructure for an AI-driven world.
• Key Use Cases with AI:
  • Economic Network: Crypto can act as "internet native money" for AI agents, which cannot use traditional banking infrastructure like credit cards.
  • Identity & Provenance: Blockchain can be used to verify if content is from a human or an AI, and to prove the authenticity of media to combat deepfakes. It provides a decentralized "registry of truth."
  • Security: The current centralized data models are "giant honeypots" for AI-powered attacks. A crypto-based public key infrastructure, using things like zero-knowledge proofs, is a much more secure architecture for an AI world.

Takeaways

• Investors should view cryptocurrency not just as a speculative asset, but as a foundational technology layer for the future of AI.
• The most promising long-term crypto investments may not be meme coins, but rather projects building the essential infrastructure for AI.
• Look for blockchain projects focused on decentralized identity (DID), data provenance, secure computing, and crypto-native payment systems for machines/AI. These are the areas where crypto provides a clear solution to problems created or amplified by AI.