AI and Digital Transformation's Impact on Investing | Global Conference 2025

Strategic "Build, Buy, and Partner" Approaches

• Consensus on Strategy: All panelists agree that organizations must utilize a hybrid "build, buy, partner, and experiment" approach rather than a single solution.
• Foundational Models: Jodi and Dennis explicitly stated their organizations will not build their own foundational models, viewing them as best left to specialized providers with proven expertise.
• Temporary Builds: Jodi noted that their firm is building solutions on a temporary basis because the AI market moves too fast for five-year decisions, with last year's decisions already requiring refactoring.
• Strategic In-House Development: Jodi cited their internal "AI research assistant" as a core strategic asset requiring in-house development to leverage specific portfolios and business processes.
• Best-of-Breed Ecosystem: Dennis emphasized that large enterprises should leverage an ecosystem of "best of breed" options to achieve outcomes rather than focusing on specific inputs or inputs.
• Private Equity Buying Focus: DJ noted that the majority of AI acquisitions in their portfolio are "buy" decisions, as they cannot compete with the $300 billion annual spend of big tech on foundational models.
• Tuck-in M&A Timing: DJ advised waiting for a "day of reckoning" where 90% of AI startups fail, making the remaining survivors viable acquisition targets for portfolio tuck-ins.
• Software Application vs. Model: Greg argued that value will accrue to companies building specific AI applications (e.g., Workday, ServiceNow) that transform value propositions, rather than those just owning underlying models.
• Valuation Premium: Jim stated there is no valuation premium for companies building their own models compared to those using third-party layers (Llama, Grok), as revenue growth is the primary driver.

Talent, Mathematics, and Infrastructure

• Mathematical Talent Scarcity: Jim highlighted the critical shortage of talent, stating that only "seven to ten true brilliant mathematicians" are needed per team to develop alpha and algorithms, noting a drastic reduction in the pool of top-tier model architects from 2,500-3,000 to roughly 250-300.
• Hiring Priority: The panel identified a bottleneck in hiring AI engineers capable of building solutions on top of models, rather than just model developers.
• Compensation Disparity: DJ observed that AI talent follows a "10x engineer" principle, where one top-tier engineer can outperform ten average ones, justifying significant salary premiums (e.g., $1M/year).
• Quantum Security Readiness: Jim warned that while quantum computers may be 7-10 years away, security protocols must be updated now to prevent "Q-Day" decryption risks to legacy 20-year-old systems.
• Talent Integration: Jim advised that AI technical leaders must report to senior levels and partner with existing IT leadership to balance rapid innovation with necessary security and stability.

AI in Asset Management and Investment Committees

• Real-Time Scenario Modeling: Dennis described the use of "agentic AI" in investment committees to process data and provide alternative perspectives in real-time, with the power of linked agents creating exponential (step-function) improvements.
• Alpha Generation: Jodi described a process of pairing AI with veteran investors to synthesize the historical writing and strategies of legends (e.g., Munger, Dalio) into current portfolio construction decisions.
• Due Diligence Acceleration: Jodi provided a case study where an analyst used AI to identify all US companies with market caps >$1B affected by tariffs within 60 seconds, a task previously impossible to update hourly.
• Dislocation Risk Assessment: DJ outlined a four-step framework for evaluating AI impact: determining if a company will be dislocated, if terminal multiples will be impacted, if operations can be improved, and if productivity can be increased.
• Private Equity Value Creation: Greg predicts the primary value in private equity will shift to helping mid-sized portfolio companies figure out their AI strategies for cost reduction and revenue growth.
• Marketing vs. Substance: DJ and Jim agreed that much of the current "AI-fying" of private equity portfolios is currently marketing theater rather than substantive operational change.
• Data Quality Warning: Jim cautioned that nine out of ten companies overrate the quality and actionability of their data, which is a prerequisite for any successful AI implementation.

Responsible AI, Trust, and Geopolitics

• Tariffs as Immediate Risk: Dennis noted a shift in conference themes, with tariffs currently ranking as the number one discussion topic over AI, though AI remains the long-term fundamental disruptor.
• Materiality-Based Governance: Jodi argued for a "materiality lens" on responsible AI, suggesting that high-risk areas like insurance claims approval require stricter guardrails than low-risk areas like lead generation.
• Misinformation and Trust: Greg and Dennis expressed deep concern regarding AI's role in spreading misinformation and the potential for "nefarious applications" to cause real-world devastation.
• Regulatory Lag: DJ warned that the proliferation of AI is moving too fast for regulators to control, creating a gap between technological capability and governance.
• Human-AI Trust Dynamics: Dennis shared an anecdote about adopting Full Self-Driving (FSD) after a positive personal experience, illustrating that trust in AI is often built through exposure and demonstrated reliability.
• Generational Trust Gaps: Jim acknowledged that trust in AI will be generational, with older demographics likely to be more skeptical of AI handling personal health or financial records.
• Deepfake Risks: Jim recounted a recent attempt to defraud his general counsel via a voice message deepfake of a board member, highlighting immediate security threats.
• Vatican Ethical Dialogue: Jim revealed he is speaking at the Vatican in June to help formulate a global, highly ethical set of rules for AI usage across sectors like medicine and finance.

Forward-Looking Statements and Predictions

• Meta-Wave Status: DJ characterized AI as a "meta wave" similar to the Internet and mobile, predicting a trajectory of fits and starts over the next 15 years rather than a straight upward line.
• IPO Market Stagnation: DJ predicted the IPO market will remain "dead" or stagnant due to geopolitical uncertainty, forcing private equity firms to sell to strategics or other PE firms.
• Consumer vs. Enterprise AI Divergence: Greg distinguished between consumer AI (personal assistants) and enterprise AI, noting that the former will drive significant disruption in e-commerce and software interactions (e.g., agents booking trips rather than humans).
• AI as Transformation Agent: Dennis predicted that AI will accelerate transformation similarly to the pandemic, where a year's worth of progress was compressed, serving as the catalyst for navigating current economic uncertainty.
• Productivity vs. Value Creation: Jodi and Greg warned that the market over-indexes on AI productivity (efficiency) gains, arguing that true value will come from revenue growth and compliance enhancements.
• Future of Private Equity: Greg suggested the future role of PE operations teams will be to help portfolio companies identify AI use cases for cost and revenue, moving beyond traditional offshoring and cost management.

Recommended AI Tools for Investors

• Jodi: Uses OpenAI (specifically version 4.0 and experimenting with reasoning models) for pre-meeting research and preparation.
• Jim: Uses OpenAI, Grok, and Gemini daily; specifically endorses "Open Evidence" for synthesizing global medical journal data for healthcare investments.
• Dennis: Uses Perplexity for search (integrated with WhatsApp) and the Aura Ring for tracking sleep patterns and health data.
• Greg: Uses OpenAI's "Deep Research" for public internet synthesis and Google's "Gemini Notebook" (private mode) for analyzing private data rooms and due diligence documents.
• Universal Recommendation: Jodi advised that turning on the Gemini function within Google Search is a critical, underutilized tool for rapid summarization and insight.