Episode Details

AI is making moves. Some are smart, some not-so-much.

Amazon just turned Alexa into an AI powerhouse, making it more than just a glorified kitchen timer. Meanwhile, AI is shaking up board game design, proving that even old-school entertainment can benefit from machine intelligence. And speaking of intelligence—XAI’s Grok3 just demonstrated how not to handle AI bias, stumbling over its so-called ‘maximally truth-seeking’ approach.

In this episode:

• OpenAI and Amazon are going all-in on voice AI
• Mercury, the new model that’s gunning for LLMs with diffusion-style thinking.
• A lawsuit from Chegg raises big questions about whether AI is killing the internet
• AI household robots

Relevant Links:

• Microsoft launches a native Copilot app for Mac with AI image and text generation.
• Open-source video AI that beats SOTA models like Sora while being 2.5x faster.
• Anthropic’s latest model with “extended thinking mode” and coding capabilities.
• Deep Research for paid users and free GPT-4o mini-powered voice.
• AI predicts protein shape changes 100,000x faster than traditional simulations.
• New humanoid robot designed for household tasks.
• AI detects flaws in board game rules pre-release.
• 80% of Americans don’t use AI at work, with mixed enthusiasm.
• xAI’s chatbot was caught censoring negative details about Musk and Trump.
• Chegg sues Google over AI-generated summaries impacting traffic.