WELCOME, EXECUTIVES AND PROFESSIONALS.

Since the last 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:

• Enterprise GenAI spending and technology evolution.

• From RAG to Autonomous Agents.

• Deloitte’s 2025 GenAI Predictions.

• Harnessing GenAI to leapfrog competitors in chemicals.

• Transformation and technology in the news.

• Career opportunities & events.

Read time: 4 minutes.

MARKET INSIGHT

Enterprise GenAI spending and technology evolution

Brief: Menlo Ventures surveyed 600 U.S. leaders for its 2024 Generative AI in the Enterprise report, highlighting key trends in enterprise GenAI spending, adoption, and technology evolution.

Breakdown:

• Enterprise GenAI spending surged to $13.8 billion in 2024, a 6x increase from $2.3 billion in 2023. Healthcare leads enterprise spending with $500 million, followed by legal ($350 million) and financial services ($100 million).

• Top use case adoption includes code generation (51%), support chatbots (31%), and enterprise search (28%). Price of tools is a minor concern (1%), with leaders prioritizing tools with potential to deliver long-term value (30%).

• OpenAI’s enterprise market share fell from 50% to 34%, while Anthropic doubled from 12% to 24%, and Google also gained traction. Enterprises now deploy three or more foundation models. Closed-source options hold 81% market share.

• RAG adoption surged to 51% in 2024, up from 31% in 2023, while only 9% of production models are fine-tuned. Agentic architectures debuted with 12% adoption, poised to disrupt the $10 trillion U.S. services economy.

• Enterprises are converging on core runtime architectures for production apps, structured across four layers: (1) compute and foundation, (2) data, (3) orchestration, and (4) tooling (detailed breakdown in full article).

• Vertical GenAI apps are gaining momentum. Initial applications focused on horizontal solutions like text and image generation, but domain-specific apps are gaining traction as enterprises seek differentiation.

Why it’s important: GenAI is set to reshape the enterprise. Spending growth, verticalization, and emerging architectures signal change ahead. Enterprises that adapt quickly to these shifts could capture outsized value.

BEST PRACTICE INSIGHT

From RAG to Autonomous Agents

Brief: Menlo Ventures article AI Agents: A New Architecture for Enterprise Automation explores six GenAI examples from RAG to Autonomous Agents, detailing reference architectures and levels of autonomy. Architecture summary here.

Breakdown:

• The fully autonomous agents of tomorrow might possess all four building blocks of AI agents: reasoning, external memory, execution, planning. But today’s LLM apps and agents do not.

• The popular RAG architecture isn’t agentic but relies on reasoning and external memory. OpenAI’s ‘Structured Outputs’ for instance enables tool use. The key distinction is that these applications use the LLM as a "tool" for search, synthesis, or generation, but the steps they take remain pre-determined by code.

• By contrast, agents emerge when the LLM controls the application's flow, dynamically deciding actions to take, tools to use, and how to interpret and respond to inputs. Menlo ventures outlines three types of agents.

• Decisioning Agent: At the most constrained end are “decisioning agent” designs, which use LLMs to traverse predefined decision trees.

• Agent on Rails: “Agents on rails” offer more freedom with a higher-level objective, but constrain the solution space with an standard operating procedure (SOP) and a predetermined library of tools to choose from.

• General AI Agents: At the far end of the spectrum are “general AI agents”, essentially for-loops with minimal data scaffolding, relying entirely on the LLMs reasoning for planning, reflection, and course correction.

Why it’s important: GenAI is entering its agents era. Understanding agentic designs is key for leveraging their potential and capturing value while managing the risks of increased autonomy.

MARKET INSIGHT

Deloitte’s 2025 GenAI Predictions

Brief: Deloitte's 2025 Predictions report highlights ten key trends, with five focused on GenAI. The insights touch on technological, societal, and business implications of GenAI's rapid evolution.

Breakdown:

• Autonomous GenAI agents are under development, with the potential to improve knowledge worker productivity and workflows. However, widespread adoption of fully autonomous systems will take time.

• As GenAI demands more power, data centers are exploring reliable, cleaner energy solutions. The tech industry should optimize infrastructure, innovate chip designs, and collaborate with energy providers to ensure a sustainable future.

• Deepfake content presents a cybersecurity-scale challenge with wide-ranging consequences. Combating fake media will escalate costs, potentially burdening consumers, creators, and advertisers with preserving internet credibility.

• The women and GenAI adoption gap is narrowing, but a trust gap persists. To unlock GenAI's full benefits for women, companies need to increase trust and reduce bias.

• On-device GenAI could make smartphones more exciting with specialized chips and extensive mobile OS integration. But the question remains: will users embrace these intelligent, next-generation capabilities?

Why it’s important: Deloitte's predictions emphasize the transformative potential of GenAI. Navigating agents, energy, deepfakes and more will define the next wave of innovation and impact.

CASE STUDY

Harnessing GenAI to leapfrog competitors in chemicals

Brief: Certain industries have been slow to adopt GenAI, despite even greater potential within their fields. McKinsey explores an example of this in its article How AI enables new possibilities in chemicals and how those who seek the initiative could leapfrog competition.

Breakdown:

• Energy and materials, including the chemicals industry, has the lowest exposure to GenAI tools at 14%, vs. 23% the cross-industry average.

• The untapped potential for GenAI comes from the industry's reliance on scientific data for innovation, fragmented customer data, and nuanced/complex manufacturing processes. Four opportunity examples below.

• Discovering new applications for existing chemicals can reduce the time to identify new applications from months to days.

• New molecule and material discovery has the potential to accelerate innovation by 2–3x, enabling the creation of new patentable chemistries and optimized products.

• Augmented knowledge extraction can improve initial manual literature assessments by 30% or more, streamlining R&D processes.

• Improving sales execution (pricing and churn management) could achieve a 2–5% increase in return on sales while reducing churn by 10–20%.

Why it’s important: GenAI presents opportunity for the chemicals industry to innovate faster, drive revenue growth, and optimize operations. Early adopters could transform their market position.

McKinsey published an article on how AI can reshape consumer experiences in healthcare, featuring five journey examples: enhancing engagement, improving quality, scheduling, follow-ups, and more.

Microsoft released a AI Strategy Roadmap. The 65-slide document covers AI readiness factors, stages of delivering AI, and industry use cases. Microsoft also refreshed its free 21-lesson course for building GenAI applications.

Bain partnered with Banca Investis to develop a GenAI app that acts as a "digital junior banker" delivering hyper-personalized banking experiences in real time. It also collaborated with Hera Group to create a roadmap of 150 GenAI opportunities, and implement solutions.

BCG unveiled a new AI Maturity Matrix assessing 73 economies on GenAI and broader AI. The 28-page report includes recommendations for driving AI adoption. Meanwhile, Booz Allen published a 48-page primer focusing on GenAI.

Sapphire Ventures, a venture capital firm, shared its framework for evaluating the future of enterprise software. Separately, Benedict Evans released his AI Eats the World report, a 90-slide analysis of market trends.

Amazon invested $4B more in Anthropic, bringing its total to $8B and strengthening their AI and cloud partnership. Meanwhile, a U.S. Congressional Commission proposed a Manhattan Project-style initiative to accelerate U.S. AI development, citing competition with China.

Microsoft introduced new specialized AI agents for Microsoft 365 at Ignite, including Copilot Actions, app development tools, and translation features. While Writer launched a self-evolving model architecture for real-time learning, improving LLM efficiency without extra training.

Google's Gemini experimental model (1121) reclaimed the top spot on the LM Arena leaderboard, the third lead change with OpenAI in a week.

Salesforce unveiled Agentforce Testing Center, a platform enabling enterprises to evaluate AI agents pre-deployment through synthetic interactions, sandbox environments, and monitoring tools.

Nvidia CEO Jensen Huang said AI hallucination issues are “several years away” from being solved, needing much more computing power. Meanwhile, an article details how Mark Zuckerberg rebuilt Meta around Llama.

CAREER OPPORTUNITIES

JPMorgan - Executive Director AI LLMs

Capgemini - AI Lead

Nordea - Responsible AI Lead

EVENTS

AWS re:Invent - December 2-6, 2024

AI Predictions 2025 - December 12, 2024

All the best,

Found this valuable? Share with a colleague.
Received this email from someone else? Sign up here
Let's connect on LinkedIn