Podcast #6

The AI landscape is shifting dramatically as of February 2026, marked by major releases and unprecedented growth metrics. OpenAI has launched GPT-5.3-Codex, a model that significantly accelerates coding tasks by merging the raw coding performance of the previous 5.2 versions with enhanced reasoning capabilities and professional knowledge. This suggests a move towards 'agentic' coding where AI takes more autonomy in software creation.

Simultaneously, Anthropic has released Claude Opus 4.6, a flagship model featuring a massive 1-million token context window even in beta. This model demonstrates state-of-the-art results in reasoning and coding, capable of handling large codebases and sustaining longer tasks without losing coherence. To prove its power, Anthropic used a team of parallel Claude instances to build a C compiler capable of compiling the Linux kernel, highlighting the potential for autonomous AI engineering teams.

On the infrastructure side, Google is reporting a staggering 52x year-over-year growth in AI usage, with its models now processing over 10 billion tokens per minute. This massive scale indicates that demand for AI compute is only just beginning. Additionally, Vercel’s v0 has evolved into a production-grade platform, allowing for secure, enterprise-level software deployment directly from AI-generated code.

Finally, the interface of how we interact with AI is changing. ElevenLabs CEO Mati Staniszewski predicts that voice will surpass text and screens to become the primary interface for AI, moving us closer to seamless, conversational interactions with technology. From massive corporate adoption to tools that automate research for just $10,000 a month, the industry is racing towards recursive self-improvement and widespread agentic workflows.