Our 234th episode with a summary and discussion of last week’s big AI news!

Recorded on 01/02/2026

Hosted by Andrey Kurenkov and Jeremie Harris

Feel free to email us your questions and feedback at contact@lastweekinai.com and/or hello@gladstone.ai

In this episode:

• Major model launches include Anthropic’s Opus 4.6 with a 1M-token context window and “agent teams,” OpenAI’s GPT-5.3 Codex and faster Codex Spark via Cerebras, and Google’s Gemini 3 Deep Think posting big jumps on ARC-AGI-2 and other STEM benchmarks amid criticism about missing safety documentation.

• Generative media advances feature ByteDance’s Seedance 2.0 text-to-video with high realism and broad prompting inputs, new image models Seedream 5.0 and Alibaba’s Qwen Image 2.0, plus xAI’s Grok Imagine API for text/image-to-video.

• Open and competitive releases expand with Zhipu’s GLM-5, DeepSeek’s 1M-token context model, Cursor Composer 1.5, and open-weight Qwen3 Coder Next using hybrid attention aimed at efficient local/agentic coding.

• Business updates include ElevenLabs raising $500M at an $11B valuation, Runway raising $315M at a $5.3B valuation, humanoid robotics firm Apptronik raising $935M at a $5.3B valuation, Waymo announcing readiness for high-volume production of its 6th-gen hardware, plus industry drama around Anthropic’s Super Bowl ad and departures from xAI.

Timestamps:

• (00:00:10) Intro / Banter

• (00:02:05) Response to listener comments

Tools & Apps

• (00:03:59) Anthropic releases Opus 4.6 with new ‘agent teams’ | TechCrunch

• (00:08:00) OpenAI’s new GPT-5.3-Codex is 25% faster and goes way beyond coding now - what’s new | ZDNET

• (00:22:02) OpenAI launches new macOS app for agentic coding | TechCrunch

• (00:23:10) Google Unveils Gemini 3 Deep Think for Science & Engineering | The Tech Buzz

• (00:27:58) ByteDance’s Seedance 2.0 Might be the Best AI Video Generator Yet - TechEBlog

• (00:31:46) China’s ByteDance, Alibaba unveil AI image tools to rival Google’s popular Nano Banana | South China Morning Post

• (00:33:26) DeepSeek boosts AI model with 10-fold token addition as Zhipu AI unveils GLM-5 | South China Morning Post

• (00:39:43) Cursor launches Composer 1.5 with upgrades for complex tasks

• (00:40:35) xAI launches Grok Imagine API for text and image to video

Applications & Business

• (00:42:19) Nvidia-backed AI voice startups ElevenLabs hits $11 billion valuation

• (00:48:36) AI video startup Runway raises $315M at $5.3B valuation, eyes more capable world models | TechCrunch

• (00:50:34) Humanoid robot startup Apptronik has now raised $935M at a $5B+ valuation | TechCrunch

• (00:53:42) Anthropic says ‘Claude will remain ad-free,’ unlike an unnamed rival | The Verge

• (00:56:50) Okay, now exactly half of xAI’s founding team has left the company | TechCrunch

• (01:00:35) Waymo’s next-gen robotaxi is ready for passengers — and also ‘high-volume production’ | The Verge

Projects & Open Source

• (01:01:31) Qwen3-Coder-Next: Pushing Small Hybrid Models on Agentic Coding

• (01:05:10) OpenClaw’s AI ‘skill’ extensions are a security nightmare | The Verge

Research & Advancements

• (01:07:12) Learning to Reason in 13 Parameters

• (01:12:33) Reinforcement World Model Learning for LLM-based Agents

• (01:16:32) Opus 4.6 on Vending-Bench – Not Just a Helpful Assistant

Policy & Safety

• (01:19:00) METR GPT-5.2

(01:23:31) The Hot Mess of AI: How Does Misalignment Scale with Model Intelligence and Task Complexity?