AI Leadership: Business Aspirations and Value Propositions | Global Conference 2025

Panel Overview & Core Theme

• Moderated by CNBC's Julia Boorstin at the Milken Institute Global Conference, the panel focused on the deployment of AI as a primary driver of economic development and business value.
• Panelists included Clara Shai (Meta Business AI), Aparna Chennapragada (Microsoft Experiences & Devices), Francis de Souza (Google Cloud), Arianna Huffington (Thrive Global), and Matt Wood (PwC).
• A consensus emerged that AI represents a unique historical inflection point, comparable to the printing press or electricity, rather than a mere incremental improvement of existing tools.

Current Adoption & Business Impact

• Meta (Clara Shai):
  • Launched "Meta AI" as a standalone app with nearly 1 billion monthly active users.
  • Llama open-source models have been downloaded over 1 billion times.
  • Strategy focuses on democratizing access, enabling small businesses (e.g., solo jewelry designers) to deploy AI agents on Instagram and WhatsApp with a single click to handle customer queries.
  • Predicts the advertising journey will evolve from static ads to interactive agent-driven conversations, potentially eliminating traditional checkout steps.
• Microsoft (Aparna Chennapragada):
  • Cites the "Work Trend Index" showing 53% of leaders want higher productivity, yet 80% of employees feel maxed out; AI is positioned to fill this capacity gap.
  • Characterizes the current shift as a compression of timelines where technology that took decades to develop now evolves in months.
  • Goal is to transition humans from "doers" to "directors," where employees manage teams of AI agents acting as world-class experts in their specific fields.
• Google Cloud (Francis de Souza):
  • Google Cloud generated $43 billion in revenue last year, a 31% year-over-year increase largely driven by AI interest.
  • Reports that over 30% of code currently in production at Google is generated by AI.
  • Note: Mark Zuckerberg stated Meta aims for 50% code generation by AI soon.
  • Google is deploying agents for complex tasks across industries, including manufacturing (Honeywell, Toyota) and media (Snap).
  • Announced a $32 billion acquisition to expand cloud security capabilities in response to AI-powered bad actors and disinformation.
• PwC (Matt Wood):
  • Over 90% of PwC's 380,000 employees are engaged in active AI projects, primarily building AI agents.
  • PwC has deployed hundreds of agents that have completed 2.2 million autonomous tasks.
  • Predicts that within the next 12 months, the primary user of AI agents will be other agents, creating an "agent-to-agent" economy where humans interact with agents on their behalf.
  • Argues that AI acts as an amplifier of human skills rather than a replacement, leading to larger, more capable teams.
• Thrive Global (Arianna Huffington):
  • Launched an AI health coach in partnership with the OpenAI Startup Fund and Eli Lilly, set to integrate with "Lilly Direct" in July.
  • Thesis: "Medicines plus healthy habits beat either alone"; AI addresses the 130 million Americans with chronic diseases to improve medication adherence.
  • Highlights that 7% of health outcomes are genetic, while habits (sleep, diet, stress) drive the remaining majority, which AI can personalize at scale.

Strategic Shifts & Workforce Implications

• The "Two-Engine" Strategy: Matt Wood advises organizations to simultaneously run a "hot" engine for innovators (safe sandboxes, rapid experimentation) and a "cold" engine for the "late majority" (clear guardrails, trust, playbooks).
• Organizational Redesign: Francis de Souza notes AI blurs traditional functional silos (sales, marketing, support), forcing a rethinking of workflows and roles where AI agents act as tools across departments.
• Human Role Evolution:
  • The workforce is shifting from executing tasks to managing, delegating, and judging the output of AI agents.
  • New skills required include "taste," "judgment," and the ability to orchestrate multiple specialized AI agents.
  • Contact centers are projected to automate routine queries while elevating higher-order, complex problem-solving to human staff.
• Stress Management: Francis de Souza notes that while automation increases efficiency, remaining human roles may be more stressful; introduces concepts like 60-second "conscious breathing" resets to manage cumulative stress.

Risks, Ethics, and Governance

• The "Five Ds" Framework (Clara Shai): Identified key risks including:
  1. Data Security (Matt Wood confirms governance accelerates adoption by reducing ambiguity).
  2. Disinformation (Francis de Souza highlights active defense against bad actors).
  3. Discrimination/Bias in model outputs.
  4. Displacement of jobs (debated as amplification vs. replacement).
  5. The Divide between those who can use AI and those who cannot.
• Human-Centric Design: Arianna Huffington warns against losing sight of the "human at the center," citing personal experiences with aging parents to emphasize that technology must serve emotional and physical needs, not just efficiency.
• Mental Health: The fastest-growing non-work use case for AI is mental health and therapy, providing a non-judgmental safe space for users to explore trauma, though concerns about long-term dependency exist.

Future Outlook (3-5 Year Predictions)

• Scientific Breakthroughs (Arianna): Prediction of a "golden age of science" where AI discovers new therapies, materials, and solutions to climate change that are currently beyond human comprehension.
• Business Model Disruption (Clara & Matt):
  • Expectation of "frontier firms" – lean, AI-ground-up companies with fewer than 10 employees generating billions in value.
  • Organizational departments may disappear or become unrecognizable; workflows will be entirely reimagined.
• Consumer Experience: The internet will evolve from human-driven navigation to an agent-driven ecosystem where humans delegate browsing, shopping, and customer support to AI agents.
• Healthcare Revolution (Arianna): Shift from "sick care" (treating illness) to proactive "health" management by addressing root causes via personalized, AI-driven lifestyle nudges.
• S-Curve Trajectory (Matt Wood): Currently in the early stages of the S-curve; rapid capability growth is expected before hitting a plateau, with the inflection point not yet reached.
• Philosophical Imperative: The technology serves as a forcing function to answer existential questions regarding human purpose, wisdom, and the definition of being human in an era of super-intelligence.

Key Disagreements & Nuances

• Replacement vs. Amplification: While the panel generally agreed on "amplification," Francis de Souza acknowledged that some roles (like code generation) will fundamentally change the economic value of software engineering, shifting value from syntax to architecture.
• Pace of Change: Clara Shai and Matt Wood suggest the pace is unprecedented compared to the cloud or mobile revolutions, which were adopted by digital natives first; AI is being adopted by massive legacy enterprises (Verizon, Toyota) immediately in production.
• Optimism vs. Realism: Clara Shai explicitly rejects doomsaying (referencing Yuval Harari), arguing that while algorithms have historically exploited human flaws, AI can now be designed to appeal to human "best" qualities (empathy, wisdom).