Ep. 59 — Hennigan's Huddle — Wednesday, May 20, 2026

Anthropic's extended thinking and agentic capabilities represent a genuine shift in AI functionality, with Claude now able to reason through multi-step problems and execute tasks autonomously in ways that would have required teams of engineers years ago. Bryan Hennigan breaks down why the next wave of AI winners will be built on reliability and trustworthiness rather than flashy demos. Topics covered: - Anthropic's agentic AI capabilities: Claude can now reason through multi-step problems, use tools autonomously, and self-check work, marking a real shift beyond marketing hype. - Realistic AI utility and limitations: Current AI tools are genuinely useful for 60-70 percent of tasks, but humans must stay in the loop for edge cases, judgment calls, and high-stakes decisions. - Privacy and data broker deception: Data brokers and AI companies are using deliberately deceptive opt-out flows and fake privacy forms, a design choice that damages industry credibility. - Safety and interpretability as competitive advantage: Anthropic's emphasis on safety over speed may look slower than OpenAI in hype cycles, but represents the right long-term business strategy. - iOS 27 focus on fundamentals: Apple is prioritizing stability, performance, and on-device AI integration rather than adding AI as marketing gloss. - App Store regulatory strategy: Apple's pre-WWDC announcements about store protections appear to be PR positioning ahead of ongoing global regulatory battles. - AEC industry data integration: Construction technology is moving toward connected data environments where design, project management, and field data seamlessly integrate across platforms.