Each week, I'll be diving into some exciting AI tools, sharing the experiments I'm working on, and offering some interesting insights about what I'm noticing in the world.

🔧 Three Tools I’m Testing

🤖 Claude 4 - Anthropic’s latest foundational models (Opus and Sonnet). I’m a big fan of their models and use them extensively for a variety of tasks. I haven’t tested the new models with coding, but they’re supposed to be impressive.

⛵Flow - Google’s new video generation tool that allows you to build multiple scenes in a larger video. I’m not much of a video editor or creator, but it’s incredible what can be created. The syncing of audio and video is impressive.

👩🏼‍🍳 Google Whisk - A tool from Google to create images that can be made from image subjects and scenes. You can then animate them with a click of a button. Fun.

🧪 AI Experiment of The Week

This week, I tried out Stitch from Google for designing a mobile app idea I have for my son. I want to create an application to help him study math by generating relevant problems tailored to his age and grade level. I intend to create this application over the summer to be ready for the start of the school year. 

I fed the new design tool this prompt to get going:

I want to create a mobile application that creates math problem sets for kids based on age and grade level. The kid should be able to answer the question by typing in the answer, writing it with their finger, or telling the application the answer. The application should keep track of how many were correct in each session. Tracking progress.

The first shot was alright, but I had a few changes in mind. I went back and forth on several things to get a design that I was okay with. However, it didn’t like one request I had for a dot grid on a canvas for drawing answers. It never completed that for me. One of the nice features is the ability to copy and paste the designs into Figma, which I tried. It worked, mostly.

Overall, it’s not a tool I’d use on anything real just yet - there’s just too many options available.

📰 Article of The Week

Replit CEO Amjad Masad claims that companies might be just months away from developing and running software without engineering teams. Speaking at a Semafor Tech event, Masad shared stories from Y Combinator startups who've been "vibe coding" their products using AI tools for three months without hiring a single developer. These founders initially planned to bring on CTOs but found Replit could fill that role instead. While Masad acknowledges we're not quite at the point where companies can run entirely without engineers, he estimates we're only 12-18 months away from that reality.

Here's where I think Masad is both right and spectacularly wrong. Yes, AI coding tools are democratizing software creation in ways that would've seemed impossible just two years ago. Non-technical founders building functional products? That's happening right now, and it's genuinely impressive.

But suggesting we won't need software developers at all? That's like saying we won't need chefs because microwave dinners exist. Sure, you can heat up a meal, but try explaining to a microwave why your dinner party needs to accommodate someone's gluten allergy, a vegan preference, and your brother-in-law's very specific opinions about seasoning.

The role of software developers isn't disappearing—it's evolving into something more sophisticated. While AI handles the routine coding grunt work, developers are becoming architects of complex systems, debugging specialists when AI goes sideways, and the crucial translators between what business stakeholders think they want and what actually needs to be built. When that AI-generated code breaks at 3 AM (and it will), you'll want someone who understands not just what the code does, but why it exists in the first place.

We're not heading toward a developer-free future. We're heading toward a future where developers focus on higher-level problem solving while AI handles the repetitive heavy lifting. The skill set is changing, but the human element remains irreplaceable.

🌎 Where the World is Going

We're witnessing the birth of a new kind of professional landscape—one where AI specialization isn't just an advantage, it's becoming as essential as knowing how to use email was in the 90s. The difference? This transformation is happening at breakneck speed, and it's creating both incredible opportunities and genuine anxiety for millions of workers.

In my corner of the world, I'm watching engineers perfect what I call "vibe coding"—that intuitive dance between human intention and AI capability where you sketch out what you want and let the machine fill in the technical details. Meanwhile, marketers are becoming prompt architects, designers are learning to collaborate with AI art tools, and even accountants are discovering how to make spreadsheets sing through natural language commands. These aren't side hustles anymore; they're core competencies that separate the thriving from the surviving.

But here's the paradox that's really striking: while we're all rushing to specialize in our AI niches, there's also this baseline of AI literacy that's becoming universal. Just like everyone eventually needed to master email and basic spreadsheets, we're all going to need fluency in communicating with AI systems, recognizing their blind spots, and knowing when to trust the machine versus our gut.

Yet I can't ignore the elephant in the room—there's a significant group of people who are genuinely terrified by this shift. They're not lazy or resistant to change; they're overwhelmed by the pace and worried about being left behind. Some are paralyzed by the fear that whatever they learn today will be obsolete tomorrow. Others are drowning in the sheer volume of new tools and techniques flooding the market weekly.

This anxiety isn't unfounded. We're asking people to fundamentally change how they work while the ground is still shifting beneath their feet. The question isn't whether AI specialization will become commonplace—it already is. The real question is whether we can bridge the gap between those racing ahead and those struggling to find their footing in this brave new world.

Specializing in AI area is going to common place. Whether that’s focusing on the areas of our day jobs, engineering, getting really good at vibe coding, or new skills. There are going to be common, general skills that everyone has to know too. 👨‍💻 About Me

🔗 Extra Links to Check Out

• Safety report on Claude 4 🦺

• Darmesh from Hubspot’s Agent.ai 🤖

• All of Google’s announcements from I/O 🎫

• Shopify launches a variety of AI features 🛒