How Amazon's #CatalogAI tackles AI quality challenges

The hardest part of AI is everything that isn’t AI. AI’s biggest roadblock isn’t the model; it’s the messy human stuff around it. In this short convo with Jim Liew, Founder of SoKat Consulting and Professor at John Hopkins University we dig into the gap between “cool demo” and “consistent business win”: data quality, incentives, governance, and who actually owns the change. If a pilot doesn’t move a number you already care about (time to value, error rate, CSAT, cost per task), it’s just homework. Curious where you’re seeing the real friction: talent, data, or adoption? #ArtificialIntelligence #AI #ProductManagement #AIOps #DataStrategy #ChangeManagement #DigitalTransformation #SMB #Leadership

🤯 Rethinking AI “Reasoning” Just read a Berkeley research paper that completely changed how I think about reasoning models. For the past year, models like o1 and R1 — the ones that “🧠 think out loud” with thousands of tokens — have been hyped as the future. The assumption? More reasoning = smarter results. But here’s the twist: researchers forced DeepSeek-R1 to skip the long reasoning chains and just give quick answers. ✅ The result: ⚡ 9x faster responses 🎯 Higher accuracy 💸 80% fewer tokens Even more surprising: running multiple short answers in parallel and picking the best consistently outperformed the long, step-by-step approach. It raises a big question: 👉 Have we been optimizing for theatre 🎭 instead of results 🚀? This might be a turning point where simpler and faster beats longer and heavier. What do you think — are long reasoning chains the future of AI, or have we been chasing the wrong signal? #AIResearch #MachineLearning #ArtificialIntelligence #Reasoning #Innovation #Productivity

🚨 95% of AI initiatives fail. MIT’s latest research is a brutal wake-up call: while most enterprises chase shiny pilots, a select few are quietly turning AI into millions in measurable value. This is the AI Divide. 👉 On one side: 95% stuck in “pilot purgatory,” bolting AI onto processes, with no learning loop and no ROI. 👉 On the other: 5% who are embedding AI into workflows, driving continuous improvement, and securing C-suite commitment. The gap is widening every quarter. Those in the 5% aren’t just using AI — they’re being transformed by it. I’ve broken down the report into a quick carousel so you can see: Why most initiatives crash and burn The 3 success patterns of the top 5% The enterprise wake-up call for leaders The question isn’t whether AI will transform your industry. It’s whether you’ll be leading that transformation — or watching from the sidelines. #AI #DigitalTransformation #CustomerExperience #KnowledgeManagement #BusinessStrategy

🌍 Globalisation is being rewritten by AI — and it’s not about geography anymore! In our latest INSEAD Tech Talk, Professors José Santos and Prof. Peter Williamson unpacked how AI is transforming the logic of international business. For over a century, companies went global through physical expansion — offices, factories, and local teams. Today, success comes from digital interaction. AI allows firms to understand what customers truly value — not from surveys, but from their behaviour. This shift demands a new strategic mindset. 💡 Key takeaways for business leaders: 🧠 Build your AI system — it’s your new competitive advantage! The edge won’t come from access to the same foundation models as everyone else, but from how you combine different layers of AI — • Deep learning to perceive and interpret data (images, language, behaviour). • Machine learning to predict and rank choices in real time. • Symbolic or rules-based AI to embed business logic, compliance, and ethics. 🚀 Increase interactions with customers. Every digital touchpoint is a learning moment. The more data you gather from real behaviour, the smarter your organisation becomes. Think of Rolls-Royce ✈️ — it no longer sells engines for airplanes, but “power by the hour.” Each flight generates live performance data that strengthens customer relationships and continuously improves the product. 🧭 Navigate the new borders. Regulation, data sovereignty, and trust are the new frontiers of global competition. The next phase of globalisation won’t be about where you operate — it will be about how intelligently you learn. 🎥 Watch the full INSEAD discussion: https://lnkd.in/dXyMYXuQ #INSEAD #AI #Strategy #GlobalBusiness #Leadership #DigitalTransformation #Innovation #TechLeadership

My key takeaways from Bits & Pretzels 🥨 First, let me share some personal insights: 1. It's totally fine to wear Lederhosen at a conference in Munich. Next time, I will wear them too! 2. Don't wait for the third day to drink your first Maß Bier 🍺; it might be canceled. Additionally my takeaways for building products with AI: ⏱️ Time in market > time to market "We're all using the same foundational models to power features. The real difference is how deeply you understand and build against your use cases. Release early and learn from usage to iterate." 🚀 Unfair advantage What is hard to replicate for others? Unfair advantages are nearly exclusively go-to-market related. 🤝 Building trust with AI products - Three essential principles: 1. Preview – Show users what the AI will do before it acts 2. Traceability – Make it clear how the AI arrived at its decisions 3. Control – Give users the ability to override and adjust AI outputs #ProductManagement #BitsAndPretzels #ProductStrategy #AI

Learnings in pursuit of AI Data structuring is often the unsung hero behind successful AI systems. Without clean, well-organised data, even the smartest algorithms struggle to deliver meaningful insights. But here’s the twist, emotional intelligence (EQ) is just as vital. Blending AI’s analytical power with human empathy can unlock new income streams for organisations ready to innovate. The real opportunity lies in aligning structured data, smart tech, and emotionally intelligent leadership. That’s where AI truly starts to pay off. #ArtificialIntelligence #DataStrategy #DigitalTransformation #Leadership

Recently, I had the opportunity to attend the Generative AI for Business short course at Melbourne Business School and it’s been an incredibly insightful experience. The program went beyond the hype to explore how AI, particularly Generative AI, is reshaping industries, decision-making, and creativity. From understanding the fundamentals of large language models to discussing ethical frameworks, data governance, and real-world case studies, the sessions provided a rich perspective on how to use AI responsibly and strategically. For me, the biggest takeaway was this: AI is here, its real and it is progressing at a rapid pace. AI isn’t about replacing people; it’s about augmenting human capability. When applied thoughtfully, it can help organisations reimagine customer engagement, improve operational efficiency, and unlock entirely new sources of value. A big thank you to the faculty at Ujwal Kayande Yalcin Akcay for an engaging and practical course, and to my fellow participants for the stimulating discussions and shared learning. The journey with AI is just beginning, and I’m excited to explore how we can responsibly turn possibility into impact. #AI #GenerativeAI #Leadership #Kearney #LifelongLearning #MelbourneBusinessSchool

🤖 3 AI Shifts I’m Paying Attention to in 2025 1️⃣ From Predictive to Prescriptive – AI doesn’t just tell us what might happen anymore; it suggests what to do next. 2️⃣ Owning Your Data & Models – depending fully on other platforms is risky. Smart teams are building their own tools. 3️⃣ Explainability = Trust – the more people understand AI decisions, the faster adoption grows. In the past few years, I’ve worked on different digital and tech projects — and one thing always stands out: 👉 Teams that start with AI in mind perform better than those that try to add it later. I’m curious — which of these 3 shifts do you think will impact your industry most in 2025? #AI #DigitalTransformation #Innovation #BusinessGrowth #TechTrends

🚀 Launching: The Human Side of GenAI Series Generative AI is transforming how we work, create, and connect — faster than any technology before it. But amid all the excitement, one truth stands out: the most powerful AI experiences are still profoundly human. Over the next few weeks, I’ll be sharing a 7-part series exploring what that really means — how we can combine human creativity, empathy, and judgment with the power of AI. Here’s what’s coming up 👇 1️⃣ Why GenAI Needs Humans More Than Ever 2️⃣ The Paradox of AI Creativity 3️⃣ The New Skill Every Professional Needs: Discernment 4️⃣ Empathy: The One Thing AI Can’t Fake 5️⃣ The Co-Pilot Mindset: Working With AI, Not For It 6️⃣ Leading with Humanity in the Age of AI 7️⃣ What Makes Us Human in the Age of Machines I’m calling it The Human Side of GenAI — a series about the technology, yes, but more importantly, about us. Follow along, join the conversation, and let’s explore how to stay human in an age of machines. #GenAI #AITransformation #HumanInTheLoop #Innovation #Leadership #AWS

📑 Conclusion What’s the biggest takeaway from our AI journey? That learning AI is not about mastering every tool—it’s about being open to experimentation, collaboration, and continuous growth. At Agora Insights, we believe that architects who embrace this mindset will lead the next wave of innovation. In this article, we reflect on: ✅ The value of learning together ✅ The power of peer support ✅ The mindset shift required to architect AI for real change AI is a journey. And it’s one we’re proud to walk with our community. 📖 Read the full reflection: https://bit.ly/4jpGgT3 #AIJourney #ArchitectingAI #BusinessArchitecture #AITraining #AgoraInsights #InnovationLeadership #FutureReady