Another interesting article I've read today. The author has been looking at an analysis of 20,000+ AI-generated applications. The conclusion is that models are actually getting better at avoiding things like SQL injection or XSS. What shows up instead is repetition — each model has its own set of common secrets it reuses across different generated apps. The same JWT signing secrets, the same placeholder passwords like password123 and admin123, appearing in app after app, along with the same endpoints. This isn’t really a coding mistake. It comes from how the model generates. Which shifts the risk. Vulnerabilities are no longer local and discovered — they become shared and predictable. Once you recognize a pattern, it is likely to appear elsewhere, as systems are generated from the same source. https://www.invicti.com/blog/security-labs/security-issues-in-vibe-coded-web-apps-analyzed?utm_source=substack&utm_medium=email

I watched an interesting podcast this weekend. This developer says AI is not a threat to junior devs — quite the opposite. Onboarding is much faster and easier, and with AI assistance they can produce initial results in weeks instead of months like before. The threat is actually for mid-level people. I didn’t fully understand why, but the junior perspective alone was eye-opening. https://www.youtube.com/watch?v=wc8FBhQtdsA

Just wanted to share some stats and reports from the MuleSoft Connect AI: Toronto, which I attended last week. If a business is still focused on digital transformation, it is already behind. The shift is underway toward the Agentic Enterprise — where deploying AI is no longer the challenge; the difficulty moves to integration, orchestration, and governance. The 2026 MuleSoft Connectivity Benchmark Report makes this visible in numbers. ✅ 94% say AI agents will require a more API-driven architecture. ✅ 96% say the success of AI agents depends on seamless, debt-free data integration. At the same time: ✅ 86% say agents introduce more complexity than value without proper integration. ✅ 64% of leaders express concern about their ability to meet near-term AI goals. ✅ Only 54% have a centralized governance framework to ensure reliability and security. Structurally, only 27% of applications are connected. Organizations operate with ~957 applications on average, rising to 1057 for those further along. Organizations keep pointing to integration as the answer. At the same time, they already observe that integration increases complexity. 𝐂𝐨𝐦𝐩𝐥𝐞𝐱𝐢𝐭𝐲 𝐢𝐬 𝐬𝐜𝐚𝐥𝐢𝐧𝐠 𝐟𝐚𝐬𝐭𝐞𝐫 𝐭𝐡𝐚𝐧 𝐜𝐨𝐧𝐭𝐫𝐨𝐥. From 2025 to 2026, integration remains a persistent issue: 95% reported it as a challenge in 2025, 82% still report it as a challenge in 2026. This reflects stabilization of a chronic condition. Agents are being connected across systems without a governing logic. 𝐈𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧 𝐛𝐞𝐜𝐨𝐦𝐞𝐬 𝐛𝐨𝐭𝐡 𝐭𝐡𝐞 𝐬𝐨𝐥𝐮𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐚𝐦𝐩𝐥𝐢𝐟𝐢𝐞𝐫 𝐨𝐟 𝐭𝐡𝐞 𝐩𝐫𝐨𝐛𝐥𝐞𝐦. I've been asked maybe 4-5 times after this event if I talk/or have any information on how to govern AI integration. Does anyone have any practices/ideas/materials on this topic? P.s. I chose it like a General Discussion, but potentially it could be a Topic Request.

After Salesforce realized that letting a few thousand employees go and trying to replace them with AI was a bad idea, they had to rehire to deal with some of the problems this entire situation created. However, no one is hiring junior-level positions. I don't think we can blame AI for this. I believe this is a sign of a struggling economy. But for us, this is a problem today and 3 years from now. I am preparing the next workshop and it seems like it will be about budgeting challenges. One area where I would typically balance a budget on a longer project is to get a mid-level developer to help with general maintenance, which after 2 months of being on a team and some good oversight, they can do. Still a net positive value add, but at lower resource cost than a senior-level dev. Well, if we are not hiring any junior-level developers today, all intermedate will become seniors and 3 years from now I will struggle to get a balanced team. And AI won’t solve this problem. What makes a good intermediate person worth bringing on the team is critical thinking, which AI struggles with. At least for now. What is the outlook in your line of work?

We alternate between leadership and management workshops. The next one is management. The last management workshop was an intro to budgeting. We could go deeper on that - or pick something completely different.