Forget Model Agnosticism: Swyx's Agent Lab Strategy
Swyx reveals the Agent Lab strategy for AI founders: own a customer's problem in a niche, become the 'AI guy,' and build deep, lasting solutions.
Podcast Crossover: AIE, AGI, frontier lab strategy with @matthew_berman and @swyxtv
With swyx, Alessio Fanelli · Sunday, July 12, 2026
In this crossover episode, Swyx, the founder of the AI Engineering Conference (AIE), discusses the origins and challenges of building the event. He delves into topics ranging from the future of AI inference chips and the performance of new frontier models like Anthropic's Fable 5, to the geopolitical implications of OpenAI's rumored equity offer to the US government and the long-term prospects of AGI alignment, emphasizing the practical strategy for founders building at the AI application layer.
Etched: Why AI's Inference Boom Demands New Chips Beyond Nvidia
Swyx explains why AI inference scale creates room for specialized chips like Etched, challenging Cerebras & Groq, not Nvidia, and why GPT-3.5 architectures make a safe bet.
LLMs Won't Deliver AGI: The Trillion-Token Trap
Latent Space's Swyx argues LLMs are hitting limits, exploring knowns not unknowns. Founders: rethink data efficiency and move beyond current pre-train/post-train paradigms.