Invest in big tech companies like Microsoft (MSFT), Google (GOOGL), and Amazon (AMZN) as a primary way to gain exposure to the AI boom through their essential cloud infrastructure. While NVIDIA (NVDA) currently dominates the AI chip market, consider diversifying into competitors like AMD as intense competition is expected to emerge over the next five years. Treat Artificial Intelligence (AI) as a core, multi-decade holding, as the technology is still in its early stages with significant growth ahead. The cryptocurrency sector is seeing renewed positive sentiment and policy tailwinds, making it worth re-evaluating for investment. Finally, monitor the rise of high-quality open-source AI models from Chinese companies like Alibaba (BABA), as they could disrupt the market.

Artificial Intelligence (AI) - Thematic Investment

• Marc Andreessen describes the current AI revolution as the biggest technological revolution of his life, larger in magnitude than the internet and comparable to the microprocessor, steam engine, or electricity.
• The current moment is seen as the culmination of an 80-year journey since the first neural network theories in the 1940s. The technology is finally delivering on its long-held promise.
• Despite the hype, the consensus is that it's still very early. The products in use today are considered rudimentary compared to what will be available in 5-10 years.
• New AI companies are showing unprecedented revenue growth, with real customer demand translating directly into dollars at a rate faster than any previous technology wave.
• The cost of AI is falling much faster than Moore's Law. The per-unit costs of all inputs are collapsing, leading to a "hyper deflation" that is expected to drive enormous demand.
• There's a debate between big models vs. small models.
  • Big Models: Often proprietary and run in large data centers (e.g., GPT-5). They will always be the "smartest" and will command a premium for tasks requiring maximum intelligence.
  • Small Models: Often open-source and can run on local hardware (like a laptop). Their capabilities are rapidly catching up to where the big models were 6-12 months prior. They are ideal for the vast number of tasks that don't require "Einstein-level" intelligence and can be deployed at a very low cost.
• The AI industry structure is expected to resemble the computer industry: a few "supercomputers" (God models) at the top, with a cascade down to a massive volume of smaller models embedded in everything.

Takeaways

• Long-Term Bullish Outlook: AI is not a short-term trend but a fundamental, multi-decade shift. Investors should consider AI a core, long-term holding in their portfolio.
• Invest Across the Stack: The opportunity isn't just in one area. Consider investments in:
  • Infrastructure providers (cloud companies, chip makers).
  • Foundation model creators (both large, closed-source and smaller, open-source players).
  • Application companies that use AI to solve specific problems in industries like law, medicine, and software development.
• Monitor the "Big vs. Small" Model Debate: The winner isn't clear, and the likely outcome is that both will thrive. This creates distinct investment opportunities. Large, centralized models benefit big tech and cloud providers, while the proliferation of small, open-source models empowers a new generation of startups and application-specific companies.

Semiconductor / Chip Industry

• The massive demand for AI has created a shortage of GPUs, which has been highly profitable for the market leader, NVIDIA.
• However, history shows that in the chip industry, "the number one cause of a glut is a shortage." The enormous profits generated by NVIDIA are acting as a "bat signal" for intense competition.
• Competition is coming from multiple angles:
  • Traditional competitors like AMD.
  • Hyperscalers (Google, Amazon, Microsoft) are building their own custom AI chips to reduce costs and dependency.
  • Chinese companies like Huawei are aggressively building their own native chip ecosystem.
  • Startups are designing entirely new chip architectures specifically for AI, which could be more efficient than the current GPUs (which were originally designed for graphics).
• It is "pretty likely" that in five years, AI chips will be "cheap and plentiful" compared to today.

Takeaways

• NVIDIA (NVDA): While NVIDIA is a fantastic company that deserves its current success, investors should be aware that its dominant position and high margins are attracting significant competition. The long-term outlook involves a more competitive and potentially lower-margin market for AI chips.
• Diversification is Key: The intense competition suggests that the value from AI compute may spread out. Investors should look beyond a single winner and consider diversifying across the semiconductor ecosystem, including competitors like AMD and the large tech companies (GOOGL, AMZN, MSFT) developing their own silicon.
• Watch for Disruptive Startups: The shift from general-purpose GPUs to specialized AI chips creates an opening for new companies. While risky, these startups could become major players or acquisition targets for larger companies.

Chinese AI Companies & Geopolitics

• China is "in the race for sure" and is one of only two major ecosystems (along with the US) building foundational AI.
• Some of the best open-source models are now coming from China, which was a surprise. Key examples mentioned:
  • Kimi: A model from a startup called Moonshot that has replicated the reasoning capabilities of GPT-5 but can run on a laptop.
  • DeepSeek: A highly capable model released by a Chinese hedge fund, which kicked off the trend of Chinese open-source AI releases.
• Major Chinese tech companies like Alibaba (BABA), Tencent, Baidu (BIDU), and ByteDance are all heavily invested in AI.
• The rise of Chinese competition has had a positive effect on US policy, reducing the appetite in Washington D.C. for "ruinous" regulations that would hinder the US's ability to compete and win the AI race.
• The cynical view is that China is "dumping" open-source models to commoditize the market and undermine Western AI companies, a strategy they've used in other industries like solar.

Takeaways

• A Two-Horse Race: The AI landscape is solidifying into a competition between the US and China. This geopolitical tension will be a major driver of innovation and government support for the industry in both countries.
• Open Source as a Disruptor: The proliferation of high-quality, free, open-source models from China could put pressure on the pricing and business models of Western companies that rely on selling access to proprietary models. This benefits companies and developers who can build on top of this free technology.
• Monitor US Policy: The competitive dynamic with China makes it less likely that the US will enact overly restrictive federal AI regulations. However, investors should watch for state-level bills (like California's proposed SB 1047), which could create a messy and harmful regulatory patchwork if not preempted by federal action.

US Big Tech & AI Startups

• Incumbents (Google, Microsoft, Amazon, Meta): These companies are "playing hard" and are central to the AI ecosystem. They are leveraging their massive cloud platforms (AWS, Azure, Google Cloud) to distribute AI "by the drink" (usage-based pricing), which has been a huge boon for startups.
• New Incumbents (OpenAI, Anthropic, xAI): These companies have established themselves as leaders in foundation models. The success of xAI, which caught up to the state-of-the-art in less than 12 months, shows that it's possible for new players to enter and compete at the highest level, suggesting no single company has a permanent lead.
• AI Application Startups: There is a "giant explosion" of startups building applications on top of foundation models.
  • The initial criticism was that these were just thin "GPT wrappers" with no defensible value.
  • However, the leading application companies (like Cursor) are evolving into deep-tech companies. They are using multiple AI models, building their own specialized small models, and backward integrating, creating a strong competitive moat.

Takeaways

• Big Tech as a Core AI Play: Investing in companies like Microsoft (MSFT), Google (GOOGL), and Amazon (AMZN) is a primary way to gain exposure to the AI boom. They benefit from both their own AI development and from selling the "picks and shovels" (cloud compute) to the entire AI ecosystem.
• Startups Are Not Squeezed Out: The idea that incumbents would dominate everything has not played out. It's proven possible for well-funded startups to catch up on foundation models, and the most vibrant area of innovation may be in the application layer, where startups can build highly valuable, specialized products.
• Pricing Model Innovation: The shift from seat-based SaaS pricing to usage-based or value-based pricing is a key trend. Companies that can price based on the value they create (e.g., a percentage of a salary saved or productivity gained) may build more valuable businesses than those simply reselling tokens.

Cryptocurrency

• Cryptocurrency mining was one of the first non-gaming applications that drove massive demand for NVIDIA's GPUs, paving the way for their use in AI.
• After a difficult "crypto winter," the industry is described as "back to being an exciting industry" due to positive policy changes.
• There is an expectation of "quite a few intersections between AI and crypto" in the future, though specific examples were not detailed.
• Many venture capital firms that "sat crypto out" may now look "a little bit less smart" as the industry rebounds.

Takeaways

• Renewed Interest: The sentiment has shifted back to positive for the crypto space. Investors who may have written off the sector could consider re-evaluating it, especially given the renewed institutional and policy tailwinds.
• Watch for AI + Crypto Convergence: This is an emerging theme. While still nascent, the combination of decentralized systems (crypto) and intelligence (AI) could unlock new types of applications and investment opportunities. Keep an eye out for projects and companies building at this intersection.