Generative AI Enterprise
Posts
The clearest picture of real-world AI

The clearest picture of real-world AI

Plus, Deloitte-Google digital twin, business model disruption, and more.

Lewis Walker
February 16, 2025

WELCOME, EXECUTIVES AND PROFESSIONALS.

Despite escalating global governance tensions, the impact of AI on work is becoming increasingly evident, compelling enterprises to reassess business models and pursue genuine transformation value.

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:

Anthropic: “The clearest picture” of AI in real-world work.
Top VC: AI agents will disrupt outsourced work.
Leaders split on priorities at global AI summit.
Deloitte & Google advance digital twins with gen AI.
Transformation and technology in the news.
Career opportunities & events.

Read time: 4 minutes.

MARKET INSIGHT

Anthropic: “The clearest picture” of AI in real-world work

Image source: Anthropic

Brief: Anthropic launched The Anthropic Economic Index revealing “the clearest picture yet” of how AI is being used in real-world work. The results, detailed in a 38-page report, show AI is primarily used to augment human capabilities, among other findings.

Breakdown:

Despite many surveys on AI’s impact on work, large-scale empirical evidence on how these systems are actually used for different tasks remains scarce.
Anthropic analyzed 4M+ Claude conversations, mapping tasks and occupations using the U.S. Department of Labor’s O*NET Database.
AI adoption is highest in software development and writing, accounting for nearly half of total use. Adoption remains early, with ~36% of occupations using AI for at least 25% of tasks.
AI is used 57% for augmentation (e.g., learning, iterating) and 43% for automation (e.g., fulfilling requests).
The true augmentation rate may be higher: e.g. a user asks Claude to draft a memo but later refines it themselves.
Anthropic says it mitigated the limitation of analyzing only Claude Free and Pro data (excluding API and Enterprise) by filtering out non-work conversations.

Why it’s important: Many AI adoption surveys exist from advisory, research, and tech firms, but Anthropic’s stands out for its scale and system-driven, empirical approach, revealing early signals of AI’s impact as it automates, augments, and expands, shaping the future of work and transformation.

BUSINESS MODEL DISRUPTION

Top VC: AI agents will disrupt outsourced work

Image source: Andreessen Horowitz

Brief: Andreessen Horowitz (a16z), a leading venture firm, explores how AI will disrupt the Business Process Outsourcing (BPO) market ($300 billion in the U.S., 2024) by in-housing largely repetitive, mundane operations with AI innovations such as voice AI agents and deep research.

Breakdown:

AI startups face incumbents, BPOs, that recognize the opportunity: Infosys deploying 100+ AI agents and Wipro reporting a 140% increase in AI adoption.
BPOs have distribution advantages, but AI startups bring radical innovation. a16z believes AI startups have the edge in building AI-native products.
BPOs often use time-and-materials pricing models, marking up labor by 20-30%. Their core business depends on people and selling their output.
Transitioning to a product-first, AI-native business model is a monumental challenge, especially for public companies. a16z argues it will compress margins, eliminate cash cows, and disrupt culture.
Leveraging cutting-edge AI is difficult with new releases coming constantly. Even experts struggle to stay updated, limiting BPOs’ adaptability.
First-class, AI-native teams stay on top of the latest AI and know how to apply it to use cases. A rare mix often absent in BPOs, a16z asserts.

Why it’s important: AI startups have a limited window to disrupt the BPO space. As foundation models stabilize and become more comprehensible to a broader audience, BPOs will increasingly integrate AI, promote in-house AI products to long-time enterprise customers, and shift to outcome-based models.

GLOBAL AI GOVERNANCE

Leaders split on priorities at global AI summit

Image source: Yahoo

Brief: At the AI Action Summit in Paris, global leaders, including politicians and CEOs, discussed AI priorities. The U.S. and UK opted out of a multinational declaration, while European leaders unveiled €200B in AI investments.

Breakdown:

Google CEO Sundar Pichai called AI "the most profound shift of our lifetimes," likening it to electricity and stressing its role in democratizing innovation.
Capgemini CEO Aiman Ezzat said EU AI regulations "went too far too fast," making it tough for global companies to deploy AI across Europe.
U.S. VP J.D. Vance warned overregulation could stifle AI growth, stating the U.S. will dominate chips, software, and global standards.
The U.S. and UK declined to sign the declaration, citing national security concerns and its "restrictive" nature, while China joined the signees.
EC President von der Leyen announced a €200B AI investment plan to make Europe an open-source alternative to global players.

Why it’s important: Despite the non-binding nature of the declaration and debates over the reality of full adherence, there is a clear lack of global consensus on AI governance. While perspectives vary, the business environment for many is shifting toward a period of lighter regulation and even faster innovation.

ACCELERATOR & CASE STUDY

Deloitte & Google advance digital twins with gen AI

Image source: Deloitte

Brief: Deloitte’s 17-page paper explores how innovations like gen AI enhance the value of digital twins in supply chain management and introduces its DT/Flash platform, built on Google Cloud, to accelerate enterprise transformation.

Breakdown:

Deloitte’s DT/Flash enables biopharma organizations to create precise digital twins of physical assets, systems, and processes in supply chain.
DT/Flash uses a universal Digital Twin Modeling Language, Digital Twin Process Logic, custom gen AI features, and graph database.
It integrates with Google’s Gemini models and Vertex AI, leveraging generative and agentic AI with multimodal capabilities.
The AI helps process unstructured data to create detailed simulations that deliver insights and enable automated optimizations.
These simulations help teams test and confirm new business decisions, deploying fixes and initiatives with confidence and accuracy.
A case study details Deloitte partnering with a biopharma company to enhance supply chain efficiency, improving inventory optimization by 20% and forecast accuracy by 15%.

Why it’s important: True transformation value comes from integrating gen AI into broader solutions. As technology advances and global dynamics shift, leveraging the latest innovation in simulation can help enterprises navigate complexity and tackle unprecedented challenges.

McKinsey outlines how AI-enabled software development cycles will shift from fragmented to AI-native, enabling faster, customer-centric, and higher-quality innovation with product managers evolving into “mini-CEOs.”

Bain explores how marketers use gen AI for campaign creation and personalization, yielding 10%-25% higher returns. Despite early success, many struggle to scale. Bain offers strategies to help CMOs drive growth.

PwC published a 17-page report on nine ways gen AI is transforming cloud adoption, from accelerating migration to enhancing application modernization, security, and compliance, highlighting real-world examples throughout.

Forrester analyst argues that the “AI Agent as Coworker” narrative is "nonsense," calling AI agents "manna from heaven" for sales tech providers, often "just automation" and exploring how the market is blurring agentic differentiation.

Protiviti’s 11-page EU AI Act guide covers implementation requirements relevant to explainability, transparency, system and risk classifications, relevant GDPR principles, and more, along with actions to ensure compliance.

Anthropic is set to launch a new AI model in the coming weeks. The model is expected to combine traditional language capabilities with advanced reasoning, excel at coding, and offer developers more control over speed vs. compute.

OpenAI CEO Sam Altman shared a roadmap update, featuring an integrated GPT-5 system that will unify the o-series and GPT-series models, that adjusts reasoning time dynamically and handles a wider range of tasks.

Thomson Reuters won a landmark AI copyright case, with the judge ruling Ross Intelligence’s use of copyrighted content for training as infringement. This could set a precedent, driving licensing deals and impacting model training economics.

Perplexity released Deep Research, a free alternative to other agentic search models. It scored 21.1% on the Humanity’s Last Exam, ranked second to OpenAI Deep Research and outperforming top models from Google and DeepSeek.

OpenAI updated its Model Spec, outlining how AI should handle controversial topics with a focus on truth, customizability, and intellectual freedom, while ensuring that guardrails remain in place to reduce the risk of real harm.