Conference Presentation, Panel, Fireside Chat
AI and Digital Transformation's Impact on Investing | Global Conference 2025
Milken InstituteDavid Steinberg, Jim Breyer, Dipanjan "DJ" Deb, Greg Emerson, Dennis Gada, Jodie Wallis, Shanali Basak
- Technology is expected to drive a profound, long-term disruption comparable to previous major booms, surpassing temporary impacts like tariffs, with AI emerging as the primary catalyst for fundamental business transformation and value creation.
- Organizations plan to adopt a "buy, partner, or leverage" strategy rather than building proprietary foundation models, opting to wait for the market to mature before investing in temporary solutions or refactoring recent decisions.
- The outlook predicts a consolidation of the underlying AI model market where only approximately five companies will emerge as dominant entities with trillion-dollar valuations, while 90% of other AI startups are expected to fail or become portfolio add-ons.
- Asset managers and private equity firms face the risk of being left behind unless they employ seven to ten elite mathematicians to develop algorithms, as human capital for architectural model development is critically scarce and more affordable than specialized engineers or physicists.
- Productivity gains are projected to be exponential when AI agents are chained together, potentially yielding a step-function improvement where four agents generate results equivalent to a thousand, enabling one engineer to become as effective as ten good engineers.
- Investments in private equity will shift focus toward helping mid-sized portfolio companies navigate AI strategy, driving cost reduction, revenue growth, and price proposition conversion, with PE firms uniquely positioned to address these needs due to their longer five to ten-year outlooks.
- Future value creation will prioritize earnings growth, revenue acceleration, and organizational efficiency over the development of internal models, as data quality is frequently overrated with nine out of ten companies failing to leverage their data effectively.
- Security and operational risks are highlighted by the potential for quantum computing in seven to ten years to decrypt existing 20-year-old protocols, necessitating immediate safety rails and the adoption of quantum algorithms to prevent bad actors from exploiting system vulnerabilities.
- The market faces a talent bottleneck for AI engineers capable of building compelling applications atop models rather than simple wrappers, alongside a critical need to re-educate the investor community to analyze assets in alignment with current AI realities.
- Enterprise AI is distinguished from consumer applications by the requirement to be native to applications rather than external additions, with specific potential to model complex interactions between tariffs, trade, taxes, and regulations in real-time for regulated industries like financial services.
- Trust mechanisms and guardrails are expected to evolve following incidents involving misinformation and deepfakes, particularly in healthcare and finance, as human trust is currently held to a lower bar for AI errors compared to human performance.
- The IPO market is anticipated to remain stagnant with conditions ranging from abundant to dead in the near term due to geopolitical issues, while software companies that transform value propositions through proprietary AI applications are expected to see massive tailwinds.
- Current AI adoption efforts in private equity are often characterized as marketing-heavy rather than substantive, creating a need to quickly validate real value to justify future valuation multiples before a value premium is recognized.
- Linear algebra remains fundamental to AI and emerging quantum algorithms, while deep tech investing involves elements of luck, with specific tools like Open Evidence being utilized as a final step in medical and healthcare investment decisions.