Every week, I embark on a journey through the fascinating realm of AI tools, unraveling the mysteries of my latest experiments. Along the way, I uncover intriguing insights and share the curious wonders I observe in the world around us.

🔧 Three Tools I’m Testing

🎥 Heygen - Video creation tool that handles complex scenarios and has a variety of AI avatars to choose from. The output is pretty incredible. I’ll be putting this to the test for my businesses over the next few weeks.

👔 Gemini 2.5 Pro (May) - Latest Gemini model. I need to come up with my own rigorous testing but the sniff test passes. It handled a variety of complex prompts well. These models are moving quickly.

🏄🏼‍♂️ Windsurf - The vibe coding IDE that I haven’t tried yet. It looks like OpenAI will be acquiring Windsurf, so we should pay attention. I’m looking forward to testing it over the next few weeks.

🧪 AI Experiment of The Week

I love baking and creating cakes. It feels like a nice escape from running tech startups. I’ve been making cakes for my kids since they were born. The last cake we made a week ago for my three-year-old was a Funfetti cake with a fondant dragon. We collaborated with ChatGPT to design the cake and then made our version.

This week, I experimented with Polymet to design a cake recipe and design application. My goal is to create an application that allows me to track cake recipes, show off bakes, and design new cakes using AI as an assistant. I’m not sure if we’ll build this out completely, but I love the idea.

The original prompt was:

Glassmorphic inspired application for managing recipes for cake

That and one more iteration led to the below design. I might feed this into Windsurf to create a working application that I can play around with. I love playing around with vibe design tools because I feel like I can describe what I’d like better than tactically implementing them.

📰 Article of The Week

UnitedHealth Group is flexing its AI muscles with over 1,000 active use cases across the organization, ranging from helping doctors take clinical notes to processing claims more efficiently. The healthcare giant is using AI for everything from predicting which patients might need extra care to automating administrative tasks, positioning itself as a leader in healthcare AI adoption. But here's the kicker – they're carefully emphasizing that AI remains a support tool, with humans firmly in control of all critical healthcare decisions, especially when it comes to approving or denying claims.

It’s interesting to watch UnitedHealth walk the tightrope between innovation and trust. They want to be seen as tech-forward (Wall Street loves that narrative), but they're also painfully aware that "AI" and "healthcare decisions" can be a PR powder keg. It's like watching someone show off their new power tools while constantly reassuring everyone they know exactly where their fingers are. The company is essentially saying, "Look at all this cool AI stuff we're doing!" while simultaneously whispering, "But don't worry, we're not letting the robots make the scary decisions." This dance reveals the reality of enterprise AI adoption – it's not the robot takeover many fear, but rather this careful, politically sensitive integration where reputation management is just as important as technological advancement.

The wider implications here are huge. UnitedHealth's approach might become the blueprint for how other healthcare companies – and really any industry dealing with sensitive decisions – handle AI integration. They're showing us that the future isn't about AI replacing human judgment but finding that sweet spot where AI handles the grunt work while humans retain authority over consequential choices. But this raises an interesting question: as AI capabilities grow exponentially, will this "humans have the final say" promise become harder to maintain? Or are we witnessing the emergence of a new corporate playbook where aggressive AI implementation comes wrapped in defensive messaging? Either way, UnitedHealth's balancing act offers a preview of how big companies will navigate the AI revolution – pushing hard on efficiency while playing it safe on trust.

🌎 Where the World is Going

The speed at which AI infiltrates every corner of our lives is staggering. One day, we ask ChatGPT for recipe ideas; the next, we trust AI doctors to diagnose our symptoms. Apps are picking our outfits based on the weather, algorithms deciding what we eat for lunch, and AI coaches crafting our workout routines. Watching this unfold, I can't shake the image of Wall-E's humans - those helpless souls floating on hover chairs, utterly dependent on machines for every basic function. Are we sleepwalking toward the Axiom?

The parallels are starting to stack up uncomfortably. We've got food delivery apps that mean we never have to leave the couch, AI that writes our emails, and cars that'll soon drive themselves. Each convenience chips away at another human skill. Remember when we could navigate without GPS? Now, I get anxious driving to a new restaurant without turn-by-turn directions, talking me through every lane change. We're outsourcing not just our labor but our thinking, our spatial awareness, and our basic problem-solving abilities.

But here's the twist that keeps me from going full doomsday prepper: humans are remarkably adaptable creatures. Throughout history, every technological leap has sparked fears of human irrelevance. The printing press was supposed to destroy memory. Calculators were going to make us mathematically illiterate. Yet here we are, using these tools to reach heights our ancestors couldn't imagine. Maybe the Axiom isn't our destination but a cautionary tale that helps us navigate this AI-powered future more thoughtfully. The key isn't avoiding the technology—it's remembering to stand up from our chairs once in a while and use our own legs.

👨‍💻 About Me

🔗 Extra Links to Check Out

• Gemini 2.5 Pro Updates ♊︎

• Anthropic Web Search API 🌐

• Don’t Lose Your Job to AI 👷🏼