How the 4 New AI Models Change How You Work
Summary
This episode analyzes four new AI models reshaping work: OpenAI’s GPT Live for natural voice interaction, Grok 4.5 for cost-efficient coding, Cognition’s SWE 1.7 for rapid software tasks, and GPT 5.6 Sol as a diligent workhorse. The host discusses how these tools shift AI from static tools to collaborative agents, highlighting trade-offs between raw intelligence, speed, and conversational fluidity.
Topics discussed
Intro and sponsor announcements
The four-day work week and new model releases
GPT Live: Full duplex architecture and voice interaction
Use cases: Language learning and consumer AI shift
Voice input vs. typing and natural conversation
AI as a cognitive presence and reasoning limits
Sponsor segments: KPMG, Robots & Pencils, Blitzy, Airtable
xAI Grok 4.5: Performance, cost, and benchmarks
Grok 4.5 vs. competitors and enterprise adoption
Moonshot KimiK 2.7 and SWE 1.7 speed breakthroughs
GPT-5.6 Sol: Diligence, legal research, and work style
Future models: GPT-5.1, GPT-6, and scaling trends
Listen ad-free on Castria