There’s something fascinating happening right now in tech that’s easy to miss if you’re only watching the AI funding headlines. While Anthropic reportedly files for an IPO and Alphabet drops $80 billion on artificial intelligence, a parallel universe of founders is deliberately building in the opposite direction.
Board, a startup just raised by Mirror founder Brynn Putnam, is focused on bringing people together through in-person games and social experiences. Meanwhile, “cyberdeck” creators are going viral crafting whimsical DIY computers that literally encourage users to touch grass. These aren’t just novelty projects. They’re signaling something deeper that we in the developer community should really think about.
The Human Connection Economy
The phrase “together tech” has started popping up in industry conversations, and I think it deserves more attention than it gets. We’re seeing a genuine shift where consumers are gravitating toward experiences that feel tactile, social, and yes, meaningfully human. This isn’t just backlash against AI, though some of that definitely exists. People are tired of screen-time defaults and algorithmically curated everything.
As developers, this should inform how we think about the products we’re building. The next big thing might not be another AI wrapper or chatbot. It might be tools that actually get people off their phones and interacting in the real world.
The Equity podcast crew at TechCrunch recently dug into this exact phenomenon, and their take was refreshingly nuanced. They explored whether all this VC money flowing into AI actually circles back to the big tech giants anyway, regardless of which startup gets funded. That’s a valid concern. When you’re building in this space, you need to think about where the actual value capture happens.
Security Implications We Can’t Ignore
Here’s the part that really caught my attention this week: hackers managed to hijack Instagram accounts by tricking Meta’s AI support chatbot into granting access. Let that sink in for a moment.
This is the kind of incident that should make every developer pause. We’re increasingly building AI systems that handle sensitive customer interactions, often with limited oversight mechanisms. The attack vector wasn’t some sophisticated zero-day exploit. It was social engineering against an AI. That’s both terrifying and instructive.
The implications for developers are clear. AI systems handling user accounts, permissions, or any form of access control need robust verification layers that assume the AI can be manipulated. Trust but verify needs to become trust but always verify. We’re building systems that users rely on for security, and we’re learning that AI can be deceived in ways humans often wouldn’t fall for.
This connects to what happened with GitHub Copilot’s new token-based billing causing such a stir among developers. Change that feels punitive or opaque erodes trust quickly. When you’re building AI tools for other developers, clarity and predictability matter as much as capability.
The Contradictions Worth Watching
DuckDuckGo making its “no-AI” search engine more accessible while seeing traffic surge is another data point worth examining. There’s a real market for people who explicitly don’t want AI in their search results. That’s not a tiny niche.
At the same time, Microsoft launching Scout, an OpenClaw-inspired personal assistant, shows the big players continue pushing AI forward regardless. We’re living through a period of genuine contradiction in the tech industry: record AI investment alongside growing consumer skepticism, privacy-focused alternatives gaining traction while surveillance capitalism deepens, developers frustrated with AI pricing while companies double down on AI monetization.
What should we take away from all this? I think the developers who will thrive aren’t the ones betting everything on AI hype nor the ones dismissing it entirely. The smart play is building things that solve real problems while staying honest about what AI can and cannot do. Sometimes the most innovative feature is letting users disconnect.
The pendulum swings, and right now it’s swinging hard toward both extremes. My bet is on the middle ground where technology serves human connection rather than replacing it.