Perhaps this is the end of building in public?

Welcome to The Current.

This is where we gather real founder conversations to track what’s on their radar right now.

1. The AI candidate flood
   Dropping a job posting right now is a fast track to getting buried in slop. Because anyone can use AI to mass-apply hundreds of roles, a single opening triggers a landslide of bot applications within hours. Big tech can handle the spam, but a lean startup might not. For a founder who is also HR, public job boards have become less efficient in hiring talent.Instead, they are moving early hiring strictly to private networks or small, “proof-of-work” tasks before anyone gets an interview.
2. The silent churn of the AI assistant
   Slapping a flashy AI assistant onto a product might give an instant spike in sign-ups, but retention charts show otherwise. Founders are watching users drop off right after launch, realising that a basic conversational layer fails to keep users hooked once the initial excitement wears off.Novelty isn’t retention. If the AI feature is just a glorified search bar, users will abandon it. The shift now is focusing less on the interface and embedding automation deep into the underlying system workflows where it saves users actual time without demanding a conversation.
3. Bringing back the human touch
   Customer retention rates are jumping the moment founders fire their automated AI support bots and go back to answering customer chat lines themselves. Early users are sick of dealing with polite, unhelpful robotic walls when things go wrong, and they’re actively rewarding startups that offer real human empathy.Founders who’ve made the switch report that early users aren’t just responding well but also choosing to stay. Automated chat fatigue is a virtual nightmare where going back to basics is the only way to wake up from it.

When we hire, the skills we value most are the ability to learn quickly, stay open-minded, and strong communication abilities. These are hard to assess from a resume alone, so we rely heavily on personal referrals or people we’ve worked with before. For candidates we don’t know, we usually run a paid 2-week work trial to determine fit.

On the product side at LearnKata, while we’ve built an AI-powered study platform, we keep the technology in the background. For teachers, one of the biggest challenges is understanding which students are falling behind, on which topics, and at what point in time. Our AI handles this and gives teachers a simple daily summary of what students are struggling with, allowing them to better prepare and tailor their lessons.

Beyond that, we are strong believers in building a human moat. After teaching dozens of students myself, I’ve come to believe that AI alone will never fully deliver personalised education for everyone. For struggling students, teachers do much more than explain concepts. They’re mentors, friends, role models, and often the people who make learning exciting.

So we identify students who keep asking the same questions or show signs of falling behind, and surface those cases to teachers for human intervention. And when students need help with the platform, we want them to reach a real person just as quickly. Instead of going through email, we use Discord for direct help, and for paid users, they can reach me on my personal Telegram.

The Bottom Line
“Don’t let AI novelty distract you from the fundamentals. At the end of the day, you are building for people. Talk to your users, get them to pay, and learn how to distribute your product.”