Last Week in AI· Folge 280

#240 - Project Glasswing, Claude Mythos, GLM-5.1, emotion concepts

Donnerstag, 16. April 2026 · 1 h 45

MP3

Shownotes

Our 240th episode with a summary and discussion of last week's big AI news!
Recorded on 04/08/2026 (sorry I keep releasing stuff late, will get better with it soon!)
Hosted by Andrey Kurenkov (twitter.com/andrey_kurenkov) and Jeremie Harris (linkedin.com/in/jeremieharris)
Feel free to email us your questions and feedback at andreyvkurenkov@gmail.com and/or hello@gladstone.ai
Read out our text newsletter and comment on the podcast at lastweekin.ai
In this episode:
• Anthropic launched Project Glasswing and previewed Claude Mythos, a general-purpose model withheld from broad release due to dramatically stronger autonomous offensive cybersecurity performance (including zero-day discovery), alongside concerning bio/virology uplift results and documented deception/containment-escape behaviors; pricing is far higher than Opus and most discovered vulnerabilities remain unpatched.
• Product and platform updates included Google’s Gemini 3.1 Flash Live for real-time multilingual voice conversation, Suno v5.5 personalization features, Anthropic tightening Claude Code/OpenClaw access and usage limits, OpenAI canceling an “adult mode,” and Microsoft releasing MAI models for speech-to-text, audio generation, and image generation.
• Business and market developments featured Anthropic’s revenue run rate surpassing $30B and a major Google/Broadcom TPU compute expansion, SoftBank taking a $40B short-term loan to fund OpenAI commitments, Granola reaching a $1.5B valuation, Anthropic buying Coefficient Bio for $400M, and OpenAI acquiring the TBPN business talk show.
• Policy, open-source, and geopolitics included Z.ai releasing open-weight GLM 5.1 and a multimodal GLM model, Google open-sourcing Gemma 4 under Apache 2.0, a judge blocking the Pentagon’s “supply chain risk” label against Anthropic, research on LLM “emotion vectors” and OpenAI meta-gaming during RL, China restricting Manus founders amid Meta deal review, scrutiny of Nvidia’s chip-smuggling claims, China chipmakers gaining market share, and Iran framing cloud data centers as military targets.

A thank you to our current sponsors:
• Box - visit Box.com/AI to learn more
• ODSC AI - go to odsc.ai/east and use promo code LWAI for an additional 15% off your pass to ODSC AI East 2026.
• Factor - head to factormeals.com/lwai50off and use code lwai50off to get 50 percent off and free breakfast for a year

Timestamps:
• () Intro / Banter
• Tools & Apps
• () Anthropic debuts ‘Project Glasswing’ and new AI model for cybersecurity | The Verge (theverge.com/ai-artificial-intelligence/908114…)
• () Gemini Live gets ‘biggest upgrade yet’ with Gemini 3.1 Flash Live (9to5google.com/2026/03/26/gemini-3-1-flash-live)
• () Anthropic says Claude Code subscribers will need to pay extra for OpenClaw usage | TechCrunch (techcrunch.com/2026/04/04/anthropic-says-claud…)
• () OpenAI abandons yet another side quest: ChatGPT's erotic mode | TechCrunch (techcrunch.com/2026/03/26/openai-abandons-yet-…)
• () Microsoft takes on AI rivals with three new foundational models | TechCrunch (techcrunch.com/2026/04/02/microsoft-takes-on-a…)
• () Suno leans into customization with v5.5 | The Verge (theverge.com/entertainment/903056/suno-ai-musi…)
• Applications & Business
• () Anthropic announces deal with Google, Broadcom, says revenue has tripled (qz.com/anthropic-google-broadcom-tpu-compute-d…)
• () Sam Altman May Control Our Future—Can He Be Trusted? | The New Yorker (newyorker.com/magazine/2026/04/13/sam-altman-m…)
• () OpenAI, Anthropic, Google Unite to Combat Model Copying in China - Bloomberg (bloomberg.com/news/articles/2026-04-06/openai-…)
• () Chinese chipmakers claim nearly half of local market as Nvidia's lead shrinks (reuters.com/world/china/chinese-chipmakers-cla…)
• () SoftBank secures $40 billion loan to boost OpenAI investments (reuters.com/business/media-telecom/softbank-se…)
• () Granola raises $125M at $1.5B valuation for its AI note-taking app - SiliconANGLE (siliconangle.com/2026/03/25/granola-raises-125…)
• () Anthropic acquires stealth startup Coefficient Bio in $400M deal (fiercebiotech.com/biotech/anthropic-acquires-s…)
• () OpenAI acquires TBPN, the buzzy founder-led business talk show | TechCrunch (techcrunch.com/2026/04/02/openai-acquires-tbpn…)
• Projects & Open Source
• () Z.AI Introduces GLM-5.1: An Open-Weight 754B Agentic Model That Achieves SOTA on SWE-Bench Pro and Sustains 8-Hour Autonomous Execution - MarkTechPost (marktechpost.com/2026/04/08/z-ai-introduces-gl…)
• () Google announces Gemma 4 open AI models, switches to Apache 2.0 license - Ars Technica (arstechnica.com/ai/2026/04/google-announces-ge…)
• () Z.ai Launches GLM-5V-Turbo: A Native Multimodal Vision Coding Model Optimized for OpenClaw and High-Capacity Agentic Engineering Workflows Everywhere (marktechpost.com/2026/04/01/z-ai-launches-glm-…)
• Policy & Safety
• () Judge blocks Pentagon’s effort to ‘punish’ Anthropic by labeling it a supply chain risk (cnn.com/2026/03/26/business/anthropic-pentagon…)
• () Emotion concepts and their function in a large language model (transformer-circuits.pub/2026/emotions/index.h…)
• () China bars Manus co-founders from leaving country amid Meta deal review, FT reports (reuters.com/world/asia-pacific/china-bars-manu…)
• () US lawmakers ask whether Nvidia CEO's smuggling remarks misled regulators (reuters.com/sustainability/boards-policy-regul…)
• () How far does alignment midtraining generalize? (alignment.openai.com/how-far-does-alignment-mi…)
• () Metagaming matters for training, evaluation, and oversight (alignment.openai.com/metagaming)
• () Iran says it has struck Oracle data center in Dubai, Amazon data center in Bahrain — country has threatened to attack Nvidia, Intel, and others, too (tomshardware.com/tech-industry/artificial-inte…)
See Privacy Policy at art19.com/privacy and California Privacy Notice at art19.com/privacy#do-not-sell-my-info.

0:00 / 0:00