Big Data and Creating Customer Growth in the Age of A.I.

Panel Composition & Core Thesis:

• David Rockafeller Jr. (Samsung Next): Believes every company must become an "AI company" to survive; emphasizes the integration of hardware and software as the future differentiator.
• Mike Flaherty (Diamond Resorts): Represents the "vacation-to-technology" shift, using big data to redefine the sales process and customer journey.
• John Shrewsbury (Wells Fargo CFO): Notes Wells Fargo spends $7–$8 billion annually on technology, currently running over 100 AI proofs of concept, with data central to serving 70 million customers.
• Mike Zuckliff (Accenture Digital): Highlights Accenture's "Applied Intelligence" group deploying 3,000 data scientists globally to solve industry-specific problems.
• Collective Stance: All panelists agree that while "General AI" is distant, "Narrow AI" is currently driving revenue, efficiency, and customer experience across all sectors.

Specific Data & Metrics:

• Diamond Resorts: Sales transformed from direct sales (90% of customers loved the product but hated the sales process) to an AI-driven experiential platform, capturing 45% of sales via the new system.
• Accenture/Toyota Pilot: An AI model predicting taxi demand in Japan resulted in a 30% revenue uplift for drivers by optimizing location positioning, reducing mileage, and increasing efficiency without the driver knowing the algorithm was active.
• Samsung Scale: Sells a TV every two seconds; smartphones operate as 200-country-connected supercomputers; billions of connected devices now exist (including 12–30 sensors per car).
• Wells Fargo: Average employee age is bimodal (concentrations at both young and older ends); 166-year-old legacy company with 10,000 physical locations, though digital interaction frequency is rising steeply.

Strategic Decisions & Business Models:

• Diamond Resorts: Launched a proprietary "10-year product" for millennials who reject perpetual commitments, moving away from the traditional "sold product" model to a "sought product" model.
• Samsung Next: Investing $12 billion annually in R&D; actively funding "picks and shovels" startups (e.g., Vicarious for robot efficiency) rather than just end-user applications.
• Wells Fargo Strategy: Moving toward "silo-breaking" data integration; customers applying for mortgages no longer need to provide repeated data as all connection points are populated; explicitly refuses to monetize customer data with third parties.
• Accenture/Toyota: Shifted marketing from "broadcasting messages" to creating "experiences," using anonymous aggregated call data to forecast events rather than targeting individuals.

Privacy, Regulation, & Ethics:

• Responsibility Framework: Accenture and Samsung prioritize "Responsible AI," requiring transparency on data sourcing, permission, and model bias before deployment.
• Regulatory Landscape: Eric Schmidt (former Google CEO) warns that strict regulations protecting a "vocal minority" (e.g., EU GDPR) may inadvertently constrain the majority and slow innovation compared to less regulated markets like China.
• China's Advantage: China is structurally advancing in medical AI by anonymizing and aggregating national medical records for research, a move US and EU regulations currently restrict.
• Customer Sentiment: A generational divide exists; younger demographics are more comfortable sharing data for convenience, while older demographics (e.g., Diamond Resorts' core 55–75 demographic) value privacy but are being forced to adapt to new "term-based" product models.
• Liability Risks: Banks face complex liability issues when using social media and chatbots (e.g., Facebook/Apple) where contractual relationships are absent.

Technology Trends & Future Outlook:

• Edge Computing: Shift from purely cloud-based processing to "edge" nodes (smartphones, cars, homes) to handle latency-sensitive tasks (e.g., autonomous vehicle swerving) and give users more granular control over data sharing.
• Data Evolution: Transition from flat, row-column data to "textured" data including 3D video, AR/VR, and streaming health metrics; requires distributed, invisible computing power.
• Human-Machine Collaboration: Consensus that machines will not replace humans entirely in the near term; machines excel at pattern recognition and anomaly detection (fraud, radiology), while humans provide context and judgment.
• Explainable AI: Industry is moving toward "OpenAI" consortia and reverse-engineering deep learning to create audit trails and explainable models for critical sectors like finance and medicine.
• Hardware Innovation: Development of neuromorphic chips designed to mimic brain organization and quantum computing to handle massive computational loads for auditing deep learning.

Disagreements & Nuanced Views:

• Moats of Big Tech: While Google/Amazon hold data moats, panelists argue these are not impenetrable due to the rapid evolution of technology and the potential for regulatory shifts to disrupt advantages.
• Adoption Rates: Debate on the speed of mobile vs. ATM adoption; mobile apps rolled out faster than physical ATM networks, yet physical branches remain vital for complex life events (retirement, real estate).
• AI Efficacy: Acknowledgment that while AI can outperform experts (e.g., radiologists), human oversight remains essential for handling "dirty, dull, or dangerous" tasks and contexts where data lacks nuance (e.g., player morale in sports).

Forward-Looking Statements:

• Wells Fargo: Predicts a 10-year future with significantly fewer physical branches, embedded finance in daily contexts, and fully automated credit/payment processes.
• Diamond Resorts: Aims to fully transition to a web-centric "sought" business model driven by big data within the next decade.
• Samsung: Envisions a future where consumers direct their own data streams (health, home, vehicle) to different services rather than relying on third-party proprietary stacks.
• Investment Strategy: Funds like WorldQuant are shifting from manual portfolio balancing to algorithmic execution and machine learning-driven investment scanning, though few let machines make final decisions unaided.