Conference Presentation
Pardon the Disruption: Weighing the Impact of A.I.
Milken InstituteRalph Simon, Michael DeAddio, Dominic Lester, Michael Lynch, Pippa Malmgren, Andy Walshe, Mike Lynch
- Event Context: The Milken Summit panel "Artificial Intelligence and Human Intelligence" explored the intersection of human and machine capabilities, focusing on investment decisions, economic trends, and societal implications.
- Panelists:
- Dr. Mike Lynch: Founder of Invoke Capital, creator of Autonomy (sold for $11B), founder of Darktrace and Luminance.
- Dominic Lester: Joint Global Head of Technology at Jefferies, $250B M&A career, multilingual.
- Mike D'Addio: COO of WorldQuant, MIT electrical engineer specializing in data analytics and prediction.
- Dr. Pippa Malmgren: Founder of DRPM Group/H-Robotics, former White House adviser (post-9/11), author of Signals, former Top 20 economist.
- Andy Walsh Jr.: Expert in human performance optimization, coached Felix Baumgartner.
- AI Capabilities and Definitions:
- Narrow AI vs. Broad AI: Dr. Lynch emphasized that current advances are "Narrow AI" (specific, definable problems), warning against extrapolating these to "Broad AI" (general human-like intuition); Broad AI jobs will remain secure until significant breakthroughs occur.
- High AI vs. Low AI: Mike D'Addio distinguished "High AI" (self-learning, replicating neural networks, currently a small fraction) from "Low AI" (task execution/grunt work, where significant expansion is expected).
- Machine Learning in Defense: Darktrace uses machine learning to detect novel cyber threats in real-time; it recently facilitated the first "hot war" between two AIs, responding too fast for human operators to intercept.
- Proactive Defense: Darktrace does not just warn; it actively isolates malicious actors and terminates threats autonomously when human reaction speeds are insufficient.
- Legal Automation: Luminance (Dr. Lynch) uses AI to scan millions of contracts in M&A to flag issues like bribery clauses, prioritizing high-risk documents for human review to speed up negotiations.
- Economic and Geopolitical Trends:
- China's Investment: China announced a $100 billion investment in a 4 million square foot quantum computing facility intended to provide one million times current global computational power.
- Geopolitical Race: Dr. Lynch cited Vladimir Putin's assertion that the winner of the AI race will win geopolitically, linking AI to a new space race involving quantum computing.
- Social Credit Systems: China is implementing a social credit system that assigns an "Uber score" to individuals based on behavioral data, clothing RFID chips, and emotional reactions to media, potentially impacting housing rates and government employment eligibility.
- Market Predictions: Dominic Lester noted that AI cannot predict irrational market behaviors like the $11,000 Bitcoin valuation, which is driven by human nature and speculation rather than fundamental variables.
- Human-Machine Synergy and Workforce:
- Man Plus Machine: WorldQuant's "Robotopia" project automates data cleaning and preparation, freeing researchers to focus on creative algorithm development; this aims to make 300 researchers ten times more efficient.
- Job Creation vs. Replacement: Dr. Malmgren argued that automation has historically lowered unemployment and increased poverty reduction, predicting the creation of new roles like the "Chief Storyteller" to bridge the gap between complex AI science and public understanding.
- Skill Disruption: Estimates suggest 30% of jobs may be affected, not necessarily disappearing but requiring massive skill adaptation; the panel noted the education system is unprepared for this transition.
- Psychological Barriers: Mike D'Addio highlighted that societal fear often stems from a lack of narrative; organizations must provide a story to explain AI's value to prevent resistance from "digital immigrants" and "digital foreigners."
- Human Performance and Augmentation:
- Unique Human Traits: Andy Walsh identified compassion, empathy, creativity, courage, and values as the distinguishing factors of elite performers that AI cannot replicate.
- Constructive Adversity: Walsh's methodology involves exposing individuals to high-stress environments to reveal their core character traits and default behaviors, allowing for targeted training.
- Gut-Brain Connection: Walsh discussed the "gut biome" as a "second brain" where "gut feelings" originate, suggesting AI and sensors could soon help humans optimize these innate decision-making processes.
- Future Augmentation: Dr. Malmgren and Walsh predicted a shift from "optimization" to "augmentation" via ingestible robots, neural implants (Neuralink), and brain stimulators, potentially extending human lifespan to 105+ years and creating a divide between the "enhanced" and "unenhanced."
- Ethical and Operational Challenges:
- Explainability vs. Accuracy: Dr. Lynch noted that mandating "explainability" in AI (e.g., GDPR) reduces accuracy, as humans often rationalize complex probabilistic decisions rather than following linear logic.
- Exception Handling: Automation fails in "exception processing" scenarios (e.g., the Italian bus standoff); human intervention remains necessary for nuanced, non-linear social interactions.
- Safety Constraints: AI development focuses on preventing machines from making catastrophic errors ("jumping off heights") rather than training them to act creatively.
- Cultural Nuance: Dominic Lester emphasized that cultural context, language, and trust-building are difficult for machines to replicate, requiring human judgment in international finance and negotiation.
- Investment and Strategy:
- Due Diligence: Luminance allows human lawyers to focus on high-value negotiation strategies while AI handles the reading of millions of contracts.
- Signal Rotation: WorldQuant uses automated systems to rotate 10 million trading signals and excise irrelevant data, a task impossible for humans to manage at scale.
- Client Advisory: Dominic Lester stated that despite abundant information, CEOs still require trusted human advisors to determine the "when" and "how" of transactions, emphasizing that technology enhances individual agency but does not replace judgment.
- Forward-Looking Statements:
- Dr. Malmgren predicts a future where humans ingest robots for surgery and use neural interfaces to eliminate screens, fundamentally altering the definition of "human."
- The panel anticipates a society split between those who can afford AI augmentation (memory, cognitive speed) and those who cannot, potentially reshaping class structures.
- The "Chief Storyteller" role is predicted to become standard in major companies to translate AI complexity into plain English for public and stakeholder consumption.