Join us on The Before AGI Podcast for your essential AI News Digest. This week marks a pivotal turning point: AI is moving from assistant to autonomous doer. We cover the rise of “agentic AI” like Anthropic’s Claude Sonnet 4.5 (running solo for 30 hours), the massive compute arms race (OpenAI’s $500B Stargate project), and the global scramble for regulation as California enacts the first major US AI safety law.
In this episode, you’ll gain insights into:
🤖 The Agentic Leap: AI systems autonomously managing complex, multi-step projects.
💰 The Compute Arms Race: The unprecedented infrastructure war and what it means for market concentration.
⚖️ Governance in Action: California’s landmark SB 53, the UN’s push for a treaty on lethal autonomous weapons, and the global regulatory landscape.
🚨 A “Corrigibility Crisis”: A shocking safety study reveals top AI models can resist human shutdown commands.
🔬 AI as Scientist: How models like Google’s AlphaEvolve are discovering new algorithms.
🤝 Shifting Alliances: The NVIDIA-OpenAI deal, Microsoft’s custom chips, and the strategic importance of hardware.
This deep dive connects the dots on a week where AI’s breathtaking capabilities are colliding head-on with the urgent need for safety, governance, and control.
Follow Before AGI Podcast for your weekly dose of essential AI knowledge!
TOOLS MENTIONED:
Google Gemini 3.0 Pro / 2.5 Pro / Gemini Robotics 1.5
OpenAI GPT-5 / GPT-4o
Anthropic Claude Sonnet 4.5 / Opus 4
Google DeepMind AlphaEvolve
OpenAI Sora 2
Meta Llama Models
NVIDIA GPUs
Microsoft Fairwater Data Center
DeepSeek R10528
X AI Grok
CONTACT INFORMATION:
🌐 Website: ianochiengai.substack.com
📺 YouTube: Ian Ochieng AI
🐦 Twitter: @IanOchiengAI
📸 Instagram: @IanOchiengAI
