Matteo Agosti

Do I actually ask for a refund? The US government decided to impose export controls on Anthropic's most powerful AI models and forcing the company to cut off access for all non-US users. Reason? Irrelevant here. More importantly, with what are we left as Europeans? Tier 2 AI products or local alternatives that are far less performant. Do we want that? Absolutely not. The AI revolution is a global one and access to AI capabilities across continents is crucial. We need to strengthen the AI ecosystem and this directive has clearly shown how vulnerable it is to political forces. Europe feels left out and this should be a wake-up call for everyone. US and Europe equally. Let's change that and work together. #AI #Sovereignty #GlobalAI #US #Europe #AIInfrastructure #nonaigenerated

61

33 Comments

Ever tried building with a headless CMS only to have your AI code assistant hallucinate non-existent API endpoints? 🤦‍♂️ Here's the problem: LLMs are trained on old data. When you're working with ButterCMS 🧈 (or any modern API-first platform), your AI assistant might suggest outdated methods, deprecated endpoints, or - worse - API calls that never existed. You end up context-switching between docs, debugging phantom code, and losing momentum. The solution? ButterCMS is now available in Context7's MCP directory. 🚀 For those unfamiliar, Context7 is a Model Context Protocol (MCP) server that pulls real-time, version-specific documentation directly into your AI coding environment. Just add "use context7" to your prompt in Cursor, Claude, or any MCP-compatible tool—and boom—you get accurate, up-to-date ButterCMS docs injected right into your context. Here's what this means for your workflow: ✅ No more API hallucinations – Get actual ButterCMS endpoints, not made-up ones ✅ Version-accurate code – Whether you're using our Blogs, Pages, or Collections endpoints, you get the right methods ✅ Faster onboarding – New devs can query "How do I fetch blog posts in ButterCMS? use context7" and get instant, accurate examples ✅ Better DX = Better docs – Great developer experience starts with documentation that's always current and accessible Use cases that just got easier: 🔹 Setting up ButterCMS with Next.js? Context7 gives you the latest integration patterns 🔹 Building custom content types? Get accurate field definitions and API responses 🔹 Implementing webhooks? See real examples from current documentation 🔹 Migrating content? Access up-to-date migration guides and API structures This is what modern developer experience looks like—when your tools understand your tools. No tab-switching. No outdated Stack Overflow answers. Just accurate, contextual help exactly when you need it. If you're building with ButterCMS (or considering a headless CMS), this changes the game. The combination of a solid API-first CMS with AI-powered, always-current documentation support means less debugging and more building. 👉 Have you tried Context7 MCP yet? What's your experience with AI coding assistants and API documentation? Drop a comment - I'd love to hear what's working (or not working) for you! Link to Docs in the First Comment #HeadlessCMS #DeveloperExperience #ButterCMS #AITools #AI #MCP #DevTools #API #DevEx #Headless #CMS #Marketer #MarTech

14

2 Comments

✨ Properly Expose your Design System to LLMs ✨ Without a tight coupling with your Design System, large-scale vibe prototypes drift from your design language pretty easily. For effective prototypes that work well in your company's context, DS compliance is a big part of the prototype experience. Here's a deep-dive on how you can markdown-ify your entire Design System (no matter how complex) so that: ✅ Your prototype starts & remains DS compliant ✅ Automatic check that runs to remove DS issues ✅ Markdown tokens remain up-to-date with original DS source 👉 Read here: https://lnkd.in/gUmUrEbG

231

12 Comments

Just read about this German startup Leapter solving one of the biggest headaches in AI development right now - the 'black box' problem. Look, we all know AI can write code faster than humans. But here's the real challenge most aren't talking about: what good is that speed if your team spends all their time trying to understand what the hell the AI actually built? Leapter's approach is smart - they're visualizing the AI-generated code through what they call 'executable models.' Instead of staring at incomprehensible blocks of code, developers can see a clear workflow of how data moves through the system. The result? Teams spending 60% less time reviewing AI-generated code. That's not just technical efficiency - it's the difference between AI being a gimmick and a genuine business advantage. I've always said the best innovations solve real problems, not imaginary ones. AI transparency isn't just a nice-to-have - it's essential if we want developers and business people to actually trust these systems. What's your experience with AI-generated code? Trust issues or smooth sailing? https://lnkd.in/e4m4hDxf

1

1 Comment

🚨 Waymo isn’t just expanding robotaxis to Munich. It is showing why regulation can become a competitive advantage. Waymo plans to bring commercial robotaxi service to Munich toward the end of 2027. 🚨 But the more interesting story started five years ago. Germany became the first EU country to establish a legal framework allowing Level 4 autonomous vehicles on public roads. Now that regulatory groundwork is helping turn Germany into a testing ground for autonomous transportation. Waymo will follow its familiar playbook: Map Munich with human drivers. Test autonomous vehicles with safety drivers. Remove the drivers. Open gradually to passengers. Then scale once regulators approve commercial operations. 💡 The technology may be moving quickly. But regulation determines where technology gets permission to scale. That distinction matters far beyond autonomous vehicles. AI is moving toward the same challenge. Companies are building increasingly capable AI agents that can make decisions, execute workflows, interact with customers, and eventually operate parts of businesses with less human supervision. The question will increasingly shift from: “Can the technology do it?” to: “Under what conditions will we allow it to?” 🎯 That creates a strategic opportunity for leaders. The organizations and countries that develop clear guardrails early may not slow innovation. They may accelerate it. Because uncertainty creates hesitation. Clear rules create a path to deployment. Germany established the regulatory road before Waymo arrived. Now Waymo has somewhere to drive. ❓The next competitive advantage in AI may not belong to whoever builds the most powerful technology. It may belong to whoever creates the clearest path for deploying it responsibly at scale. #AI #ArtificialIntelligence #AILeadership #DigitalTransformation #Innovation #FutureOfMobility

47

2 Comments

The open source AI space is moving fast and GTC Berlin has a full track dedicated to it. Open models, open ecosystems, the tooling and frameworks that the developer community is building on: it's all there. And it's not just talk. Partners like Black Forest Labs are presenting live, and the conversations in this track are as close to the frontier as it gets. If you build on open models, contribute to open source AI projects, or think the democratisation of AI capability matters, this is what you can’t miss this GTC: Agentic AI in Production: From Open Models to Agents You Own [SB61091] Shopify Fine-Tunes Open Source Models and Optimizes Inference Cost [SB61184] Building Open Models That Developers Can Inspect, Adapt, and Deploy [SB61090] Connect with the Experts: Open Model Recipes [CWEB1056] #NVIDIAGTC Berlin #OpenSource #OpenModels #NVIDIA #AI

2