Last Week in AI #242 - ChatGPT Images 2.0, Qwen 3.6 Max, Kimi-K2.6
136 snips
Apr 29, 2026 They review breakthroughs in image models that generate crisp, readable text and screenshot-like results. They cover major Chinese and open model releases, including a trillion-parameter MoE and new API-only offerings. They discuss research tools that automate deep investigations and how AI is being used to find and fix hundreds of software bugs. They also highlight big cloud investments, chip and startup moves, and rising AI-generated media spam.
AI Snips
Chapters
Transcript
Episode notes
Qwen 3.6 Max Marks Monetization Of Chinese Model Network
- Alibaba's Qwen 3.6 Max Preview moved to API-only with shut down of the free tier, signaling a shift from community growth to monetization.
- Jeremie points out Qwen overtook Llama as most self-hosted globally, then Alibaba monetized that network effect.
LLMs Enabled A Massive One-Time Security Overhaul For Firefox
- Mozilla used Anthropic's Mythos to find and fix 271 Firefox bugs, showing LLMs can accelerate large-scale security audits.
- Jeremie and Andrey discuss this as a one-time overhaul effect and question whether software security becomes an ongoing escalation or reaches near-perfect reliability.
Treat Early App Integrations As Experiments
- Expect first-generation app integrations (like Starbucks in ChatGPT) to be clunky and iterate toward native parity over time.
- Andrey recommends treating these as early experiments, not replacements for vendor apps until UX and flows are refined.