This time we celebrate my new AI podcast. I wresteled with doing this because I had a podcast ( several podcasts ) before and it's a struggle to be consistent. But when you keep getting confirmations, you just need to challenge yourself and Go! I listened to @Taneisha Fair who said my voice needs to be present in this space. The news this week is brought to you by High Tech Tea. 📖 Episode Summary Episode 1 covers everything that dropped in AI over the past 7–10 days. Two stories out of China, one from OpenAI, and one spicy, creative-world debate. The through-line: AI is getting sharper, not easier, and sharp tools require sharp operators. Stories covered this episode: 1. Kimi K2.6 by Moonshot AI — the open-source trillion-parameter model you need to know 2. OpenAI Codex + Chronicle — screen-aware AI that watches, learns, and shows up ready 3. Claude as a Design Tool — Figma killer or prettier slop? Jeneba's hot take Hot Topics: - Google Personal Intelligence — they already have your data. Now they want to make it useful. - Canva AI 2.0 — full prompt-based overhaul, and it did not ask for permission - Claude Opus 4.7 — Anthropic's sharpest model yet (but it'll drain your limits fast) 🍵 THE DROP: New Tools That Slapped This Week (Or Are Coming Soon) 🔧 Tool What It Does Kimi K2.6 — Moonshot AI Open-source, 1 trillion-parameter multimodal model. 300 sub-agents running 4,000 parallel workflows + 13 hours of autonomous coding. #1 on Humanity's Last Exam. - OpenAI Codex + Chronicle Screen-aware coding agent inside ChatGPT. Reads your screen in real time, learns your stack, and shows up already warmed up — no re-explaining required. - Claude Design — Anthropic Generates UI, design systems, decks, and PowerPoints at the intent layer spacing, hierarchy, accessibility, IA. Not a Figma killer (yet), but a serious first-draft engine. - Google Personal Intelligence Gemini + AI Mode in Search now plug into Gmail, Photos, and Google apps for hyper-personalized responses. Free in the US, personal accounts only ,Workspace is next. - Canva AI 2.0 Full prompt-based platform overhaul. Persistent brand memory, object-based editing, HTML imports in Canva Code, and Slack/Gmail/Drive/Calendar connectors. Research preview. Claude Opus 4.7 Anthropic's most precise model yet. Self-verification, long-horizon coherence, 3× better vision, new "Extra High" effort tier. Stops fixing weak prompts for you.