Episode Details

Monday’s show used Super Bowl AI advertising as a starting point to examine the widening gap between AI hype and real-world usage. The discussion moved from ads and wearable AI into hands-on model performance, agent workflows, and recent research on reasoning models that internally debate and self-correct. The throughline was clear, AI capability is advancing quickly, but adoption, trust, and everyday use continue to lag far behind.

Key Points Discussed

00:00:00 👋 Opening, Monday post–Super Bowl framing

00:01:25 📺 Super Bowl ad costs and AI’s visibility during the broadcast

00:04:10 🧠 Anthropic’s Super Bowl messaging and positioning

00:07:05 🕶️ Meta smart glasses, sports use cases, and real-world risk

00:11:45 ⚖️ AI vs crypto comparisons, hype cycles and false parallels

00:16:30 📈 Why AI differs from crypto as a productivity technology

00:20:20 📰 Sam Altman media comments and model timing speculation

00:24:10 🧑‍💻 Codex hands-on experience, autonomy strengths and failure modes

00:29:10 📊 Claude vs Codex for spreadsheets and office workflows

00:34:00 💳 GenSpark credits and experimentation incentives

00:37:10 💻 Rabbit Cyber Deck announcement and portable “vibe coding”

00:41:20 🗣️ Ambient AI behavior, Alexa whispering incident, trust boundaries

00:46:10 🎥 The Thinking Game documentary and DeepMind history

00:49:40 🧠 David Silver leaves DeepMind, Ineffable Intelligence launch

00:53:10 🔬 Axiom Math solving unsolved problems with AI

00:56:10 🧠 Reasoning models, internal debate, and “societies of thought” research

00:58:30 🏁 Wrap-up, adoption gap, and closing remarks

The Daily AI Show Co Hosts: Beth Lyons, Andy Halliday, and Karl Yeh