Michael Eisenberg: How China Could Overtake the US in the AI Race | E1167

Intro (00:00:00)

• Foundation models are the fastest appreciating asset in history.
• Lena Connell openly stated that AI is a threat to American capitalism.
• The IPO window is wide open, but the question is what price one is willing to take.
• Michael Eisenberg and Lex Fridman discuss the potential of China overtaking the US in the AI race.
• China has a number of advantages over the US in the AI race, including:
  • A larger population and a larger market for AI applications.
  • A more centralized government that can make decisions more quickly and efficiently.
  • A culture that is more accepting of risk and experimentation.
• The US has a number of advantages over China in the AI race, including:
  • A more established AI research community.
  • A more robust venture capital ecosystem.
  • A stronger intellectual property protection regime.
• The outcome of the AI race between China and the US is uncertain, but it is clear that China is a serious contender.

Tech Revolution or Market Mania? (00:00:56)

• AI is the most transformative technology of our time, but there's a significant financial gold rush surrounding it.
• Bubbles create necessary infrastructure for future advancements, but many investors will lose money in the process.
• Some investors are overly exposed to the same companies through multiple funds, while others chase logos and replicate successful models without considering local factors.
• Foundation models appreciate rapidly, but financial markets don't handle minute-to-minute updates well.
• The accuracy of AI training models is affected by rapid information changes, especially in finance and stocks.
• The value of AI companies lies in the teams that can bring models together, not in the models themselves.
• AI models depreciate quickly, and talent retention is challenging, posing financial risks to AI companies.
• The speaker doubts the long-term viability of many AI companies due to the lack of lasting assets and high talent turnover.

AI Showdown Between Tech Titans & Innovators (00:06:31)

• The future of AI lies in integrating it into people's daily lives, such as using AI screens to fetch personalized articles from the New York Times.
• Foundation models are not a lucrative investment for venture capitalists, as only a few companies will make significant profits while the majority will incur losses.
• The emergence of large language models (LLMs) since 2013 represents a significant moment in AI.
• Israel lacks the talent density for foundation models due to its focus on military technology rather than academia, while France excels in AI due to its strong academic foundation.
• Instead of solely investing in foundation models, the focus should be on applied AI and fully integrating AI systems.
• AI will create a divide between "AI countries" and "non-AI countries," as well as "AI companies" and "non-AI companies."
• Unlike with the internet, legacy companies will find it harder to catch up with AI companies because they need to have their data set up correctly to benefit from AI.
• Customer service AI and AI SDR for sales teams are competitive areas with many competitors.
• More interesting investment opportunities lie in areas where specific domain knowledge is required, making it harder for others to compete.

Investing in Competitive Markets (00:12:49)

• Market education is preferable to competing in saturated markets, as it allows companies to create new demand and establish a unique position.
• Synthetic biology and chemistry rocket fuel companies offer a unique value proposition by providing synthetically engineered rocket fuel at a lower cost, despite being in a competitive industry.
• SAS businesses are inherently riskier than venture businesses due to intense competition and price erosion.
• Companies that offer a unique way to play a future trend in the public markets tend to receive premium multiples, as they provide investors with exclusive access to that trend.
• Horizontal SAS processes will be disrupted by AI.

Horizontal SaaS vs. Vertical (00:17:39)

• Most successful companies today are horizontal SaaS companies (e.g., Airtable, Notion).
• The rumor is that Amazon has advised against buying software as AI can handle most business process needs.
• More companies will likely build their own software or use tools to create unique business processing software.
• Accenture's success in generative AI revenue is due to companies lacking the knowledge to integrate AI into their business units.
• McKenzie is also doing well in these at-risk fee businesses, offering consulting services to improve clients' bottom and top lines using software and data.
• The market for purchasing software is slowing down as companies have become saturated with basic software and are looking for new revenue opportunities.
• Buyers today are willing to pay for value, and consulting firms like Accenture and McKenzie are leading the way in this trend.
• Corporate buyers are more interested in new revenue opportunities rather than incremental efficiency gains from software.
• The challenge lies in efficiently extracting value from providing millions in savings to customers, as traditional SaaS models may not capture the full value provided.

Adoption Cycles & Technological Growth (00:21:52)

• Adoption of AI is happening rapidly, but its broad economic impact will take longer than expected.
• Regulators may impose punitive policies or regulations on data access due to fear of losing control, leading to a plateau in AI development.
• Europe is more likely to face regulatory challenges than the US, which could hinder its competitiveness in AI.
• The US, China, and Israel are the key markets for AI development and are likely to have less regulation compared to Europe, giving them an advantage.

The Role of AI in National Defense (00:24:22)

• AI is likely to be a key competitive advantage in wars.
• Closing the kill chain using AI is being used in Israel and by European armies.
• The 21st century is about AI countries versus non-AI countries.
• Iran is a threat as they are looking to acquire nuclear weapons and have missiles that can hit Europe.
• Keeping AI systems closed is difficult as there are ways to bypass restrictions.
• The US needs to be ever more competitive technologically to stay ahead of adversaries.
• China is 10 years ahead of the US in using AI to manipulate the minds of young people through TikTok.
• It is important to assume that competitors are smarter and stay ahead.

The Ethical Dilemma of Defense Investment (00:27:19)

• Michael Eisenberg cautions against the recent enthusiasm for defense investments, emphasizing the seriousness of funding efficient killing and the maturity of Israeli kids who have served in battles.
• He highlights the alliance between Silicon Valley and the defense industry, stressing patriotism and the importance of respecting the gravity of defense technology sales.
• Eisenberg warns investors about pursuing unfamiliar territories in defense and hard tech without proper knowledge, predicting losses for software investors.
• He suggests that great founders from non-software backgrounds may not be interested in SAS investors due to the inapplicability of the SAS playbook to harder technologies.
• Eisenberg discusses advancements in synthetic biology, its increasing accessibility, and his personal journey of learning chemistry through AI tools and expert collaboration.
• He emphasizes the importance of humility, acknowledging limitations, and seeking expertise from others, as demonstrated in his collaboration to build a synthetic biology company.
• Eisenberg mentions his investment in empathy.com and the value of diverse perspectives in building successful companies.

Concerns Over Liquidity: IPOs, M&A, and IP (00:34:41)

• The majority of companies (98%) face challenges in obtaining liquidity.
• Companies considering an initial public offering (IPO) may need to accept lower prices due to current market conditions.
• It is advisable for companies to go public early and continue their growth in the public markets.
• Private equity (PE) firms are likely to acquire a small percentage of companies, be sensitive to pricing, and offer lower multiples due to higher interest rates.
• Concerns exist regarding the potential impact of artificial intelligence (AI) on software-as-a-service (SaaS) companies and their valuations.
• Some PE funds that recently experienced significant write-downs on software purchases may have a reduced appetite for further investments in the sector.
• While exceptional businesses can emerge from venture capital, achieving moderate returns on medium-sized companies can still be a positive outcome.
• Liquidity remains a crucial factor, and secondary markets may not always provide genuine buyers.

Selling Positions: All or Incremental? (00:41:03)

• Michael Eisenberg advocates for a binary approach to venture capital investing, either taking a whole position or none at all, and advises against selling all of a position, suggesting instead to keep a portion to ride the upside.
• He emphasizes the importance of having an information advantage and acting on convictions, criticizing the strategy of placing multiple bets and concentrating capital into winners, as the best-performing companies often have suitors and are difficult to invest more in, while bad ones should not receive additional funding.
• Eisenberg highlights the value of investing in middling companies with potential, as these can sometimes turn into big winners with additional capital.
• He acknowledges the difference between consumer and enterprise investments, with consumer investments being more transient and unpredictable, while enterprise investments offer more predictability.
• Venture capitalists often underestimate the value of moats and competitive advantages in businesses, leading to insufficient focus on deepening these advantages.
• The sustainability of competitive advantages depends on consumer behavior and the potential shift to different modalities beyond search.
• Only a small number of companies within a portfolio truly matter in terms of driving growth and sustainability.

From Hyperscalers to Sustainable Growth (00:49:57)

• The tech industry is transitioning from rapid growth to slower, sustainable growth, posing challenges for companies with moderate growth rates.
• Michael Eisenberg reflects on improving his communication style with entrepreneurs, aiming to be less direct and more caring.
• Eisenberg suggests shortening board meetings to 45 minutes, focusing on core strategic issues, and prefers in-person meetings for better engagement.
• He emphasizes the importance of meeting entrepreneurs in person to avoid excessive pontificating and maintaining a high-level perspective to avoid overwhelming them.
• Eisenberg highlights the potential of a company building a unique monetization layer for LLMs using blockchain, founded during the pandemic.
• He favors founders who are new to a market (net new naive) and advises against relying too heavily on expert opinions, as they can sometimes hinder progress.

Quick-Fire Round (00:56:32)

• Younger Israelis are challenging the stereotype of being overly absorbed in social media by actively fighting for important causes.
• Despite attracting significant foreign investment, the Israeli startup ecosystem needs help from foreign investors to scale its businesses.
• Venture capitalist Michael Eisenberg believes in respecting competitors and constantly pushing himself to work harder.
• Eisenberg warns of the risks of the zero-interest rate environment, particularly the potential for governments to seize assets.
• Despite data suggesting a decline in performance, Eisenberg aims to continue contributing to the content platform business in the next ten years.
• Eisenberg relies on referrals from his extensive experience rather than proactive sourcing for deals.
• While acknowledging the need to improve in providing management help, Eisenberg excels in networking and values his global network.
• Eisenberg believes in the importance of networking and reaching out to anyone in the world.
• He acknowledges that his time management may be perceived as unbalanced, but he suggests asking others for their perspective.
• Eisenberg expresses gratitude for the opportunity to have an in-person conversation and appreciates being accepted as an LP in Harry's fund.