Generative AI Enterprise
Posts
What Harvard learned from 776 professionals

What Harvard learned from 776 professionals

Plus, Deloitte-NVIDIA agents, OpenAI case study, and more.

Lewis Walker
March 23, 2025

WELCOME, EXECUTIVES AND PROFESSIONALS.

Harvard's study with Procter & Gamble uncovers how AI reshapes not only performance but also expertise and social connectivity in teams, compelling organizations to rethink the very structure of collaborative work.

Since the previous edition, we've reviewed hundreds of the latest insights on best practices, case studies, and innovation. Here’s the top 1%...

In today’s edition:

Harvard insights from 776 professionals.
Accenture: Reinventing enterprise operating models.
Stanford: The artificially intelligent boardroom.
OpenAI and Booking.com personalize travel at scale.
Deloitte launches Zora AI agents with NVIDIA.
Transformation and technology in the news.
Career opportunities & events.

Read time: 4 minutes.

MARKET INSIGHT

Harvard insights from 776 professionals

Image source: Harvard Business School

Brief: Harvard Business School conducted a study with 776 professionals at consumer goods giant Procter & Gamble (P&G) to assess AI’s impact on individuals and teams in developing products and retail strategies.

Breakdown:

Individuals randomly assigned to use AI, GPT-4 and GPT-4o in this study, performed as well as a team of two without AI.
Teams using AI were far more likely to produce top-tier solutions, ranking in the top 10% for quality, as shown in the graph above.
Both AI-enabled individuals and teams worked faster, saving 12-16% of the time spent by those not using AI.
AI broke down silos, enabling R&D teams to be more business-driven and commercial teams to create more technical solutions.
Teams using AI were happier, reporting higher energy and enthusiasm, while also experiencing less anxiety and frustration than non-AI users.

Why it’s important: In recent years, much of the focus has been on productivity of individual knowledge workers like consultants, lawyers, and coders. But a lot of work happens in teams, which offer benefits individuals alone can't typically provide, such as better performance and satisfaction.

WORKFORCE EVOLUTION

Accenture: Reinventing enterprise operating models

Image source: Accenture

Brief: Accenture’s 34-slide report explores the future of enterprise operating models and organizational design through four lenses: amplified intelligence, dynamic skills, fluid boundaries, and adaptable structures.

Breakdown:

Gen AI agents act as “intelligent colleagues,” enhancing collaboration, decision-making, and collective intelligence.
As the nature of work evolves, enterprises need to shift from traditional, predefined career paths to more dynamic models.
Gen AI accelerates information flow, breaking down hierarchies and silos to enable cross-disciplinary work.
Organizational structures will flatten, enabling self-organizing teams and requiring new governance, metrics, and funding approaches.
Accenture’s Marketing & Comms case study shows how organizations can reinvent the workforce while increasing brand value by 25%.

Why it’s important: Across industries and regions, executives are moving from piloting targeted gen AI use cases to scaling these technologies. However, in the push for ROI, often too little attention is given to how gen AI will fundamentally reshape enterprise structure and ways of working.

AI GOVERNANCE

Stanford: The artificially intelligent boardroom

Image source: Stanford Hoover Institution

Brief: Stanford's Hoover Institution released a 5-page brief detailing how AI will reduce information asymmetry between management and directors, and alter the process by which boards fulfil their governance obligations.

Breakdown:

AI will enhance scenario planning, risk analysis, and investment prioritization in-house, reducing reliance on consultants and cutting costs.
Improve compensation committee pay benchmarking, predict proxy advisor recommendations, and assess tax/legal impacts faster.
Enhance audit committee fraud surveillance, internal controls, reasonableness checks, and balance automation vs. human oversight.
Identify workforce skill gaps and forecast capacity needs. Monitor legal and regulatory changes, including lawsuits and enforcement actions.
Track and assess board effectiveness in time allocation, focus, and balancing reactive vs. proactive decision-making.

Why it’s important: While boards recognize the vast potential of AI, it's important to consider how it can reshape board operations and practices itself, with the prospect of substantially improving corporate governance quality in how boards function, process information and interact with management.

CASE STUDY

OpenAI and Booking.com personalize travel at scale

Image source: OpenAI

Brief: OpenAI shared how Booking.com, one of the world’s largest travel platforms, integrated OpenAI’s large language models (LLMs) to enable smarter search, faster support, and personalized, intent-driven travel experiences.

Breakdown:

Booking.com had used machine learning for a decade, but they struggled to capture user intent, particularly during the early discovering phase.
Despite hundreds of search filters, they only helped if travelers knew exactly what to look for. LLMs enabled a more conversational experience.
The AI Trip Planner let users ask natural language questions like, “Where should I go for a romantic weekend in Europe?”
The first prototype, which responded with destinations, built itineraries, real-time pricing and availability, was launched in 10 weeks.
Success led to smart AI filters, property Q&A, AI review summaries, and "help me reply", increasing engagement and bookings.

Why it’s important: New technology can unlock previously unviable opportunities, and build momentum for further innovation. Booking.com's Senior Director, Adrienne Enggist, said they now aim to "build a concierge-like companion" to guide travelers throughout their entire journey.

ACCELERATOR

Deloitte launches Zora AI agents with NVIDIA

Image source: Deloitte

Brief: Deloitte, in partnership with NVIDIA, launched Zora AI™, a suit of agents that perceive, reason and act to autonomously execute tasks with speed and precision, improving human productivity and decision-making.

Breakdown:

The platform leverages NVIDIA’s new Llama Nemotron reasoning models and AI-Q Blueprint for developing agentic systems.
Early benchmarks show the Llama Nemotron 'Super' version outperforms Llama 3.3 and DeepSeek R1 in STEM and tool testing.
Zora AI has agents optimized for an expanding portfolio of use cases in finance, HR, supply chain, procurement, sales, and more.
Hewlett Packard Enterprise (HPE) is using Zora AI in Finance for financial statement analysis, scenario modeling, and market analysis.
Deloitte is using Zora AI for Finance internally, automating expense management, cutting costs by 25%, and increasing productivity by 40%.

Why it’s important: Professional services firms like Deloitte, Accenture, EY, and others are rapidly integrating agentic technologies into new or existing offerings. With the right partnerships and expertise to optimize technology and drive adoption, enterprises can unlock new value realization opportunities.

IBM in partnership with Adobe and AWS, published a 23-page report, on the state and opportunities of organizations’ content supply chains (CSCs).

PwC shared 22-slides on how boards can oversee AI, covering strategy, governance, regulations, and talent to drive responsible and valuable AI use.

Cognizant’s AI adoption report highlights leadership commitment but business readiness gaps. 70% of leaders fear they’re not moving fast enough on gen AI.

McKinsey shared insights on how COOs can maximize AI’s impact, from rethinking operations to strengthening COO-CIO collaboration.

BCG’s CEO workout for peak AI performance explores improving focus, risk management, resilience, and buy-in to drive AI impact.

Accenture shared the front-runner's guide to scaling AI, revealing that 34% of firms have scaled at least one "strategic bet."

EY explores how gen AI can transform tourism through personalized interactions, and improved customer communication.

Andreessen Horowitz outlines how Model-Context-Protocol (MCP) impacts AI-tool interactions, key market players and and what’s being built.

NVIDIA’s GTC 2025, kicked off by Jensen Huang's keynote, revealed new open reasoning models, chip releases, robotics, autonomous vehicles, and more.

Adobe launched its AI agent strategy, introducing the Experience Platform Agent Orchestrator and ten agents for enterprise tasks in marketing, and more.

Baidu released ultra-cost-effective AI models: ERNIE 4.5 reportedly at 1% the cost of GPT 4.5, and ERNIE X1 matching DeepSeek's R1 at half the price.

OpenAI launched its next-gen API-based models for text-to-speech and speech-to-text, enabling custom AI voices and improved speech recognition.

Anthropic added web search to Claude, giving it access to real-time data and closing the feature gap with competitors like ChatGPT and Gemini.

Google introduced Canvas in Gemini, a collaborative space for document editing and code creation, along with the addition of Audio Overviews.

Meta announced that its Llama open-source model reached 1B downloads this month, up from 650M in December 2024.

Y Combinator reported that a quarter of their startups now have 95% of their code written by AI, helping teams reach $10M in revenue faster than ever.