Daniel Spurling

AI Transformation isn’t just about the tech. It’s about alignment. Capability. Context. But here’s the catch: most orgs skip the part that makes it all work. That missing piece is called AI Enablement—and it’s not just support. It’s the function that turns strategy into reality. I just published a new post on what AI Enablement is, why it matters, and how it anchors real transformation. Read here 👉 https://lnkd.in/exfXnyZg Does your org treat enablement like a function—or an afterthought? If this interests you, let’s have a discussion. What are your experiences with AI Enablement and Transformation? #AIStrategy #AIEnablement #Innovation #AI

4

2 Comments

For hundreds of thousands of years, humans got individually smarter through better tools, primitive symbolic communication, and basic planning. But innovation was siloed with each innovator and disappeared with them. Then, around 70,000 years ago, something changed. Humans developed the ability to perform three specific things: (1) share intent (2) build cumulative knowledge (3) reason collectively. This cognitive revolution launched exponential progress that continues today. 𝐓𝐡𝐢𝐬 𝐞𝐱𝐚𝐜𝐭 𝐭𝐫𝐚𝐣𝐞𝐜𝐭𝐨𝐫𝐲 𝐢𝐬 𝐫𝐞𝐩𝐥𝐚𝐲𝐢𝐧𝐠 𝐨𝐮𝐭 𝐢𝐧 𝐬𝐢𝐥𝐢𝐜𝐨𝐧. Today’s AI agents are increasingly capable and powerful. They can connect with each other, but they can’t think together. The next shift is to enable an agentic cognitive evolution. It requires infrastructure that allows agents and humans to align on intent, synchronize cognition state, context and knowledge, and innovate collectively. Intelligence compounds collectively. I wrote an architectural blueprint for this shift, what we call the 𝐈𝐧𝐭𝐞𝐫𝐧𝐞𝐭 𝐨𝐟 𝐂𝐨𝐠𝐧𝐢𝐭𝐢𝐨𝐧. It introduces new mechanisms for (1) semantic state coordination, (2) persistent collective memory, and (3) shared reasoning with guardrails. Together, these form the foundation for 𝐝𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐞𝐝 𝐚𝐫𝐭𝐢𝐟𝐢𝐜𝐢𝐚𝐥 𝐬𝐮𝐩𝐞𝐫𝐢𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞. The same transformation that language brought to human society, engineered for silicon. 👇  Link in comments to the blog (and white paper), 𝐹𝑟𝑜𝑚 𝐶𝑜𝑛𝑛𝑒𝑐𝑡𝑖𝑜𝑛 𝑡𝑜 𝐶𝑜𝑔𝑛𝑖𝑡𝑖𝑜𝑛: 𝑆𝑐𝑎𝑙𝑖𝑛𝑔 𝑂𝑢𝑡 𝑆𝑢𝑝𝑒𝑟𝑖𝑛𝑡𝑒𝑙𝑙𝑖𝑔𝑒𝑛𝑐𝑒

555

78 Comments

Which Tech Skills are Losing Value Quietly? The rapid adoption of Generative AI and automated workflows in 2026 has fundamentally shifted the "technical baseline." Skills that once commanded high salaries are now being commoditized as AI matures from a novelty into a core infrastructure tool. This analysis explores the tech skills losing market value—not because they are useless, but because they are no longer "rare." We examine the transition from execution-based roles to judgment-based roles, specifically within the context of the Nigerian education system and the global digital economy. #TechTrends2026 #FutureOfWork #CareerAdvice #DigitalTransformation #AIImpact #TechLayoffs

1

Bravo to Anthropic for their "Keep Thinking" Campaign. https://lnkd.in/gukmjfZW When we do an AI44 Readiness Assessment among leaders trying to implement AI in their companies, the biggest barriers aren't cost, accuracy, or security. It’s this specific fear: “What if AI makes my team think less?” 😱 Based on what we see in the workplace, that fear is justified. There is a real risk that employees who are told to “leverage AI” without further coaching or direction will use it to avoid thinking and churn out AI slop instead of higher quality thought. But it doesn’t have to be that way. If you show your employees how to use generative AI to "think more, sooner, and better" (as we would say at Twenty44), generative AI can become a tool to augment your thinking and produce more valuable output. That’s why Anthropic’s new “Keep Thinking” message is a good one. Instead of promising to think for you (like most other generative AI promotion I see), it reminds us: Don’t stop thinking, use generative AI to "Keep Thinking" and think better. Also, LOVE the music track. 👉 How are you helping your teams avoid the “AI slop” trap and use Gen AI to think more, sooner and better? twenty44 #GenerativeAI #AIAdoption #FutureOfWork #AIInTheWorkplace #ThinkMoreSoonerBetter #ArtificialIntelligence #KeepThinking #AIReadinessAssessment #AI44ReadinessAssessment

10

1 Comment

🚀 Beyond Basic RAG: Agentic RAG If you're building with LLMs, you know Retrieval-Augmented Generation (RAG) was a game-changer for grounding models in fact. But as we push for more complex solutions, its "retrieve-then-generate" limitations are showing. It's time to evolve from a static retriever to a dynamic problem-solver. 🧱 Traditional RAG: The Stateless Retriever Standard RAG follows a simple, linear process: Retrieve – Conduct a vector search to fetch top-k relevant chunks. Augment – Append retrieved data to the original prompt. Generate – The LLM creates a response based on this combined context. 🔍 But it’s limited by design: ❌ No feedback loop – Poor retrieval leads to poor output; there's no chance to course-correct. ❌ Single-pass logic – It can’t merge or reason across diverse sources (e.g., comparing SQL data with PDF reports). ❌ Shallow understanding – It matches on surface-level semantics without truly grasping user intent. 🧠 Agentic RAG: The Stateful Orchestrator Agentic RAG introduces a paradigm shift: from passive retrieval to active reasoning and orchestration. It embeds a smart agent that mirrors how a human analyst would work. Here’s how it works: 🧩 Query Decomposition – Breaks down complex prompts into actionable sub-tasks. 🛠 Dynamic Tool Selection – Uses the best tool for each task: vector DB, SQL, REST API, file search, web queries—even other AI agents. 🔁 Iterative Reasoning – Assesses intermediate results, reflects, and refines its strategy as needed. 🧪 Validated Generation – Synthesizes answers only after gathering and verifying the necessary information from multiple sources. Agentic RAG doesn’t just answer—it thinks, strategizes, gathers, and then responds. 🌟 Why Agentic RAG Matters This evolution unlocks a new wave of AI-powered systems that go beyond basic Q&A: ✨ Intelligent Copilots – Assistants that reason and respond across steps. 🕵️ Research Agents – Deep-dive investigators that synthesize from disparate formats. 🌐 Multi-Modal Orchestrators – Integrate insights from text, tables, APIs, and more. 🤖 Autonomous Knowledge Workers – Agents that handle complexity, ambiguity, and dynamic workflows. The Bottom Line: ▶️ RAG answers a question based on retrieved data. ▶️ Agentic RAG first understands the question, creates a strategy, uses multiple tools to gather comprehensive data, and then solves the problem. This shift is what will take us from simple Q&A bots to AI systems that can execute complex, multi-step tasks autonomously. 💬 Let’s Talk What do you think are the biggest roadblocks to adopting Agentic RAG in real-world production systems? Are we ready to shift from “retrieve and respond” to “understand and solve”? #AgenticRAG #RAG #LLM #AIagents #GenerativeAI #MachineLearning #AIengineering #ArtificialIntelligence #TechLeadership #FutureOfAI #CognitiveAI

136

3 Comments

Most teams don’t struggle with AI agents because the technology is immature. They struggle because prompting is not treated with design discipline. An agent prompt is not a single instruction. It’s a system of assumptions, constraints, and decision rules. When those aren’t explicit, reliability breaks down fast. That’s why two teams can use the same models and get completely different outcomes. This guide breaks down how to design agent prompts the way operators design systems. Intentionally, repeatably, and with scale in mind. Full breakdown here: https://lnkd.in/gzhuQgRW #Agents #Prompting #AI

295

12 Comments

I've been sending around this HBR article (in the comments) on why transformation stalls at the "last mile". The article identifies seven structural frictions blocking AI transformation: productivity gaps, tribal knowledge, architectural complexity, and governance. What was underweighted for me in the list of frictions: change management and risk management. The article frames the identity problem as a knowledge capture challenge. How do you get what's in people's heads into the system? But there's a deeper issue. When you ask someone who has built their career around being the person who knows to hand that knowledge to a machine, you're asking them to redefine their professional identity. That requires trust, time, and intentional change management. The same is true for risk. The governance section is framed operationally: things like who can create agents and where accountability sits. But organizations scaling AI are also taking on new categories of risk: model bias, decision transparency, regulatory exposure, and reputational harm. Those deserve to be first-class considerations in any transformation. Great technology, poorly managed, creates great problems at scale. #AI #DigitalTransformation #ChangeManagement #Leadership #RiskManagement #Experient

36

2 Comments

Why does AI adoption stall even when the technology works? Last November at Deloitte CONNECT, I spent time with leaders, from Olympic gold medalists to tech CEOs, discussing how AI is reshaping leadership, talent, and trust. The recurring theme: adoption falters when people don’t trust how systems behave or how they’re governed. I joined Ashley Reichheld in a conversation captured by WIRED to talk about what it takes to turn trust from theory to practice, and how we’re doing that at Autodesk through our AI Transparency Cards: https://lnkd.in/gTkCBceP

68

2 Comments

__________ An unexpected dilema for AI ________ Computing states, "growing number of company bosses are questioning the business case for AI after a global survey found most have yet to see financial benefits from the technology. The study, conducted by PwC, surveyed 4,454 business leaders across 95 countries and territories." For just two factors, these numbers guarantee statistical accuracy. If you do not believe me, ask AI. “It found that 56% of CEOs reported seeing neither increased revenue nor reduced costs from their use of AI, despite what the firms described as "significant" levels of investment. Only 12% of respondents said AI had delivered both lower costs and higher revenues". Why is this so? - IMO, it is because billions of investments and enormous efforts of BigTech moguls have had different putposes. They do not warry about your ROI but concern about your subordinative behaviour and following the directions they dictate. The deeper your business adopts AI they create, the more obidient you’ll be. https://lnkd.in/ecAFs4PK

1

The systems integration industry just died!! Most consulting firms don't know it yet. For forty years, the model was simple: complexity is inevitable, hire an army of humans to manage it, bill by the hour. We called it "digital manual labor." Agentic AI just destroyed that overnight. When agents move data between systems with near-zero friction, the entire business of selling "integration" evaporates. If you're in consulting, delete these three things immediately: 1. Delete the Bench A deep bench of junior developers isn't an asset. It's dead weight. Pyramid structures are collapsing. Diamond structures are emerging. 2. Delete the "Implementation Phase" Build and Test used to take 70% of budgets and months of time. Agents don't need that infrastructure. Implementation isn't a phase anymore. It's an event. Minutes, not quarters. 3. Delete the Billable Hour If an agent audits a ledger in 30 seconds, charging by the hour is suicide. You can't sell the input (time). You must sell the outcome (value created). The winners? The Creators!!! Not the implementers. The people who see patterns no one else sees and design solutions that don't exist yet. The ones who charge for outcomes and value added—not hours logged. We're witnessing the Great Decoupling: process from software, execution from value. The companies that win won't have the most developers. They'll have the best pattern recognizers orchestrating agents to deliver outcomes. So here's the disruption: Stop paying for the bench. Start paying for the brain! Stop buying implementation. Start buying outcomes! Stop charging by the hour. Start charging for value! The question for every consulting firm, dev shop, and services company: Are you still selling the process—or are you finally selling the result? Drop a comment: If you're in consulting or services—what are you seeing? Are firms adapting or pretending this isn't happening? If you're a founder—are you still paying for hours or demanding outcomes? Because one model has a future. And one doesn't.........

3

1 Comment