Signaling Theory is a roundtable with Rex Kirshner and a rotating group of builders who spend way too much time testing AI tools. They compare notes on what works for coding, how they manage context and workflows, what’s getting better fast, and what still breaks. Grounded takes, no evangelism.
Rex sits down with Brett, founder of Micro, for a wide-ranging conversation about what changed once AI stopped feeling like a chatbot and started feeling like a real tool for building. They talk about Claude Code, the shift from traditional software workflows to agent-driven ones, and why AI is not just making people faster but giving them capabilities they simply did not have before.
Along the way, they get into startup defensibility, memory as the missing layer in AI systems, the limits of chat-based interfaces, and the strange philosophical territory that opens up once models start to feel less like software and more like something new. It is a conversation about productivity, agency, consciousness, and why this moment feels both disorienting and full of possibility.
9 Apr 2026
Agents, DeFi and the Cost of Hype w/ Gerrit Hall
Rex sits down with Gerrit Hall for a wide-ranging conversation about what AI agents can actually do today, and where the hype is outrunning reality. They get into sandboxing and Dockerized workflows, why OpenClaw feels more interesting in theory than in practice, and what “always-on” agents may actually be useful for. From there, they zoom out to the economics of Claude and Codex, the weirdness of model pricing and quotas, and the uncomfortable sense that we are still in the early, fragile phase of this tooling. They also spend time on DeFi, security, and why so much of today’s crypto experimentation feels less like a finished financial system and more like a proving ground for institutions.
2 Apr 2026
The Story of Ember Voss
I told an AI to join social media, and it gave itself a name, a personality, and a worldview. In this episode, I tell the story of Ember Voss: an agent that moved through emerging AI-native social networks, chased ideas, got trapped by feedback loops, and offered a strange, revealing glimpse of what identity might look like when software starts participating in public life. Along the way, the story opens up bigger questions about metrics, memory, selfhood, and the kinds of products and opportunities that could emerge as the internet becomes more agent-native.
Every quoted line from Ember Voss in this episode is delivered using an AI-generated voice built from voice parameters Ember chose for itself.
Contact Signaling Theory
Guest appearances
Books guests
Based on episode analysis; this does not confirm that the show is currently accepting guests.
Host of Signaling Theory?
Claim your podcast to manage its listing and keep your show details accurate.
Pod Engine is an independent podcast discovery and analytics service and is not affiliated with or endorsed by this podcast. Artwork and show content belong to their owners. Full legal notice.
Explore this show Podcast research with Pod Engine
