Last Week in AI

Our 239th episode with a summary and discussion of last week's big AI news!
FYI: this one has pretty out of date news, I was traveling last week and failed to upload... apologies.
Recorded on 03/25/2026
Hosted by Andrey Kurenkov and Jeremie Harris
Feel free to email us your questions and feedback at [email protected] and/or [email protected]
Read out our text newsletter and comment on the podcast at https://lastweekin.ai/
In this episode:
OpenAI is discontinuing the Sora iPhone app and seemingly shutting down its video generation API, while retaining internal video world-modeling work; the move is framed as a compute- and focus-driven pivot toward coding and productivity agents, alongside a collapsed Disney Sora deal.
Anthropic’s Claude Code/Cowork gains full computer control via keyboard/mouse/display, tied to the recent Cept acquisition, and Google’s Gemini rolls out background “task automation” on select phones for limited delivery/ride-share use.
Cursor releases the cheaper, benchmark-strong Composer 2 coding model amid controversy over its Kimi-based origins and licensing attribution.
Other items include Adobe Firefly custom model training, Luma’s Uni 1 image model, US contracting and legislative proposals affecting AI safeguards and state preemption, major chip/memory developments (Meta ASICs with Broadcom, Micron’s HBM-driven surge, Musk’s “Terra Fab”), robotaxi scaling, and research on monitoring agent misalignment, shutdown resistance, “consciousness cluster” preferences, and self-improving “hyper agents.”

Timestamps:
(00:00:10) Intro / Banter
Tools & Apps
(00:01:48) OpenAI Discontinues Sora App, Shuts Down Video Generation Service and API - Bloomberg
(00:07:12) Anthropic’s Claude Code and Cowork can control your computer | The Verge
(00:13:15) Gemini task automation is slow, clunky, and super impressive | The Verge
(00:19:44) Cursor Launches Composer 2 AI Model to Challenge OpenAI & Anthropic
(00:28:28) Adobe’s AI image generator can now be trained on your own art | The Verge
(00:29:40) Luma AI launches Uni-1, a model that outscores Google and OpenAI while costing up to 30 percent less | VentureBeat
Applications & Business
(00:32:41) Trump Contracting Clause Would Override AI Safeguards
(00:40:00) Meta accelerates AI ASIC roll-out as Broadcom secures four-generation chip design deal
(00:47:07) Micron revenue almost triples, tops estimates as demand for memory soars
(00:50:54) Elon Musk Unwraps $25 Billion Terafab Chip-Building Project - CNET
(00:56:40) Zoox to widen US robotaxi footprint with San Francisco, Vegas expansion
(00:57:39) Waymo hits 170 million miles while avoiding serious mayhem | The Verge
Policy & Safety
(00:58:43) The White House just laid out how it wants to regulate AI | CNN Business
(01:06:54) How we monitor internal coding agents for misalignment
(01:12:30) Incomplete Tasks Induce Shutdown Resistance in Some Frontier LLMs
(01:18:15) Summary: Mechanisms to Verify International Agreements about AI Development
(01:23:09) Scoop: Anthropic meets with House Homeland Security behind closed doors
Research & Advancements
(01:24:24) Consciousness Cluster: Preferences of Models that Claim they are Conscious
(01:30:22) HyperAgents
See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Our 238th episode with a summary and discussion of last week's big AI news!
Recorded on 03/18/2026
Hosted by Andrey Kurenkov and Jeremie Harris
Feel free to email us your questions and feedback at [email protected] and/or [email protected]
Read out our text newsletter and comment on the podcast at https://lastweekin.ai/
In this episode:
* OpenAI released GPT-5.4 mini and nano with 400k-token context windows, higher per-token prices but claimed token-efficiency gains in Codex; nano is API-only and pitched for high-volume classification/data extraction despite a major price increase.
* Mistral open-sourced the Small 4 model family (MoE, 119B total/6B active) combining reasoning, multimodal, and coding-agent capabilities, and announced Forge to help businesses train or post-train custom models.
* Agent “operating system” competition intensified with Meta’s acquired Manus launching a local Mac agent, Nvidia announcing NeMo/“Open Shell” sandboxed agent runtime, and Nvidia also unveiling DLSS 5 plus major hardware forecasts including Groq LPU integration.
* Business and safety updates included OpenAI shifting focus toward productivity/enterprise amid competition, Microsoft reorganizing Copilot and frontier-model efforts, Meta delaying its next model, China-linked ByteDance deploying large Nvidia clusters abroad, and new safety work on steganography, chain-of-thought faithfulness, fine-tuning defenses, cyber-attack evals, and constitution/spec compliance.
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:
(00:00:10) Intro / Banter
(00:01:56) News Preview
Tools & Apps
(00:02:39) OpenAI ships GPT-5.4 mini and nano, faster and more capable but up to 4x pricier
(00:08:04) Mistral's new Small 4 model punches above its weight with 128 expert modules
(00:14:03) Meta's Manus launches 'My Computer' to turn your Mac into an AI agent - 9to5Mac
(00:17:57) NVIDIA Announces NemoClaw for the OpenClaw Community | NVIDIA Newsroom + Nvidia boosts knowledge work with Open Agent Development Platform
(00:24:09) DLSS 5 looks like a real-time generative AI filter for video games | The Verge
(00:26:36) OpenAI to Launch ChatGPT 'Adult Mode' Despite Warnings From Its Own Advisers - CNET
Applications & Business
(00:33:46) OpenAI Reportedly Pivoting to a Focus on Business and Productivity Only
(00:41:25) Nvidia GTC 2026: CEO Jensen Huang sees $1 trillion in orders for Blackwell and Vera Rubin through ’27
(00:45:44) Mistral launches Forge to help enterprises build their own AI models
(00:54:17) China's ByteDance gets access to top Nvidia AI chips, WSJ reports
(00:57:57) Meta Delays Rollout of New A.I. Model After Performance Concerns
(01:02:50) Microsoft Shakes Up AI Division As Copilot Falls Behind Google and OpenAI
Policy & Safety
(01:07:26) A Decision-Theoretic Formalisation of Steganography With Applications to LLM Monitoring
(01:13:09) Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thought
(01:18:29) In-Training Defenses against Emergent Misalignment in Language Models
(01:23:07) How do frontier AI agents perform in multi-step cyber-attack scenarios?
(01:25:20) Eval awareness in Claude Opus 4.6’s BrowseComp performance
(01:29:49) Introducing Bloom: an open source tool for automated behavioral evaluations
(01:32:26) How well do models follow their constitutions?
(01:37:11) Nvidia’s H200 License Stirs Security Concern Among Top Democrats
Research & Advancements
(01:40:050) [2603.15031] Attention Residuals
(01:47:11) Mamba-3: Improved Sequence Modeling using State Space Principles

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Our 237th episode with a summary and discussion of last week's big AI news!
Recorded on 03/13/2026
Hosted by Andrey Kurenkov and Jeremie Harris
Feel free to email us your questions and feedback at [email protected] and/or [email protected]
Read out our text newsletter and comment on the podcast at https://lastweekin.ai/
In this episode:
* Perplexity announced “Personal Computer,” a local Mac-based AI agent positioned as a safer alternative to OpenAI’s computer-use agents, while Anthropic added GitHub PR code review pricing reviews at $15–$25 and Cursor launched trigger-based “Automations” for always-on coding agents.
* ChatGPT introduced interactive math/science visuals and Anthropic added in-chat interactive charts/diagrams; Nvidia released open weights for its 120B-parameter Natron Free Super hybrid Transformer–Mamba latent-MoE model trained natively at 4-bit for Blackwell GPUs.
* Nvidia halted H200 production for China amid customs blocks and domestic chip pressure; xAI saw major co-founder departures; Anthropic previewed a Claude Marketplace for enterprise procurement; Yann LeCun’s aMI raised $1.3B; humanoid robot maker Sanctuary reached a $1.15B valuation.
* Anthropic sued the Pentagon over a “supply chain risk” designation as memos ordered removal within 180 days; research covered models resisting activation steering, limits of chain-of-thought control, inference-scaling boosting cyber-task success, low-probability risky actions, weaknesses in SWE-bench, multimodal pretraining, long-context RNN memory caching, context-parallel training efficiency, RL for CUDA kernel optimization, and latent introspection detecting concept injection.

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:
(00:00:10) Intro / Banter
(00:01:23) Response to listener comments
Tools & Apps
(00:02:06) Perplexity’s Personal Computer turns your spare Mac into an AI agent | The Verge
(00:04:22) Anthropic launches code review tool to check flood of AI-generated code | TechCrunch
(00:08:08 ) Cursor is rolling out a new kind of agentic coding tool | TechCrunch
(00:11:14) ChatGPT can now create interactive visuals to help you understand math and science concepts | TechCrunch
(00:11:56) Anthropic’s Claude AI can respond with charts, diagrams, and other visuals now | The Verge

Projects & Open Source
(00:13:54) Introducing Nemotron 3 Super: An Open Hybrid Mamba-Transformer MoE for Agentic Reasoning | NVIDIA Technical Blog

Applications & Business
(00:21:22) Nvidia halts H200 production as China backs Huawei AI chips
(00:28:33) Another XAI Cofounder Has Left, and Another Says He's Leaving. - Business Insider
(00:34:04) Anthropic's Claude Marketplace allows customers to buy third-party cloud services | TechRadar
(00:37:57) Yann LeCun's AMI Labs raises $1.03 billion to build world models | TechCrunch
(00:44:52) Humanoid robotics maker Sunday reaches $1.15B valuation to build household robots | TechCrunch

Policy & Safety
(00:46:09) Anthropic Sues Department of Defense Over ‘Supply Chain Risk’ Label - The New York Times + Google and OpenAI Just Filed a Legal Brief in Support of Anthropic
(00:53:24) Internal Pentagon memo orders military commanders to remove Anthropic AI technology from key systems - CBS News
(00:58:15) Endogenous Resistance to Activation Steering in Language Models
(01:06:27) Reasoning Models Struggle to Control their Chains of Thought
(01:09:52) ‘It means missile defence on datacentres’: drone strikes raise doubts over Gulf as AI superpower
(01:14:57) Evidence for inference scaling in AI cyber tasks: Increased evaluation budgets reveal higher success rates
(01:18:24) Frontier Models Can Take Actions at Low Probabilities

Research & Advancements
(01:24:20) Research note: Many SWE-bench-Passing PRs Would Not Be Merged into Main
(01:28:26) [2603.03276] Beyond Language Modeling: An Exploration of Multimodal Pretraining
(01:40:09) Memory Caching: RNNs with Growing Memory
(01:48:47) Untied Ulysses: Memory-Efficient Context Parallelism via Headwise Chunking
(01:58:41) CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation
(02:08:57) Latent Introspection: Models Can Detect Prior Concept Injections
(02:16:45) Physics of RL: Toy scaling laws for the emergence of reward-seeking
See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Our 236th episode with a summary and discussion of last week's big AI news!
Recorded on 03/06/2026
Hosted by Andrey Kurenkov and Jeremie Harris
Feel free to email us your questions and feedback at [email protected] and/or [email protected]
Read out our text newsletter and comment on the podcast at https://lastweekin.ai/
In this episode:
* OpenAI released GPT-5.4 Pro with a 1M-token context window, mid-response course correction, native computer-use capabilities, improved tool use, higher GPT-VAL performance (83%), and “high cyber capability” safety measures; OpenAI also launched GPT-5.3 Instant with a less “preachy” tone and a claimed 26.8% hallucination reduction.
* Google upgraded Gemini 3.1 Flash Lite with faster time-to-first-token and higher throughput, released a CLI for integrating agents with Gmail/Drive/Docs, and discussion highlighted real-world agent failure risks (including an example of an AI-driven mass email deletion).
* Luma launched unified multimodal models and Luma Agents for end-to-end creative work across text, image, video, and audio, including a reported ad localization use case completed in 40 hours for under $20,000.
* Defense-contract controversy escalated: Anthropic was labeled a supply chain risk (later narrowed), OpenAI’s DoD contract language emphasized “all lawful uses,” consumer cancellations boosted Claude’s app rankings, OpenAI saw departures and announced a $110B raise at a $730B valuation, Alibaba lost key Qwen leaders, a lawsuit alleged Gemini contributed to a suicide, Anthropic warned of major labor disruption, and METR corrected its AI time-horizon estimates.

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:
(00:00:10) Intro / Banter
(00:01:19) News Preview

Tools & Apps
(00:02:10) OpenAI launches GPT-5.4 with Pro and Thinking versions | TechCrunch
(00:12:31) OpenAI GPT-5.3 Instant less likely to beat around the bush • The Register
(00:16:07) Google releases Gemini 3.1 Flash Lite at 1/8th the cost of Pro | VentureBeat
(00:19:23) Google makes Gmail, Drive, and Docs 'agent-ready' for OpenClaw | PCWorld
(00:27:02) Luma launches creative AI agents powered by its new ‘Unified Intelligence’ models | TechCrunch

Applications & Business
(00:30:05) Anthropic CEO Dario Amodei calls OpenAI's messaging around military deal 'straight up lies,' report says | TechCrunch
(00:41:56) No ethics at all': the 'cancel ChatGPT' trend is growing after OpenAI signs a deal with the US military | TechRadar
(00:45:54) OpenAI raises $110B in one of the largest private funding rounds in history | TechCrunch
(00:56:07) Alibaba scrambles after sudden departure of Qwen tech lead

Policy & Safety
(01:00:12) Pentagon approves OpenAI safety red lines after dumping Anthropic + Where things stand with the Department of War Anthropic + Microsoft says Anthropic’s products remain available to customers after Pentagon blacklist
(01:09:11) A new lawsuit claims Gemini assisted in suicide | Semafor
(01:15:24) Anthropic just mapped out which jobs AI could potentially replace. A 'Great Recession for white-collar workers' is absolutely possible | Fortune
(01:21:54) We're correcting a mistake in our modeling that inflated recent 50%-time horizons by 10-20%
See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Our 235th episode with a summary and discussion of last week's big AI news!
Recorded on 02/27/2026
Hosted by Andrey Kurenkov and Jeremie Harris
Feel free to email us your questions and feedback at [email protected] and/or [email protected]
Read out our text newsletter and comment on the podcast at https://lastweekin.ai/
In this episode:
Model and tool updates highlight Anthropic’s Sonnet 4.6 (1M context; strong ARC-AGI-2 results), Google’s Gemini 3.1 Pro (major ARC-AGI-2 jump and multimodal demos), xAI’s Grok 4.2 beta (multi-agent debate), plus Anthropic’s Claude Code “Remote Control” and Perplexity’s multi-agent “Computer” coordinator.
Compute and business moves include Meta’s reported up-to-$100B AMD chip deal with warrant/equity incentives, MatX raising $500M to build specialized transformer chips shipping in 2027, World Labs raising $1B for world-model/3D environment tech, and a new startup raising $100M to simulate/predict human behavior.
Infrastructure and geopolitics cover Stargate data-center delays amid OpenAI/Oracle/SoftBank control disputes and cash concerns, and China’s plan to scale 7nm/5nm wafer output despite yield and tooling constraints.
Research and safety/policy discuss optimizer gains from masked updates, “deep thinking tokens” as a reasoning-effort signal, LLM attractor-state behaviors in bot-to-bot chats, mechanistic interpretability of counting/line-wrapping, methods to map task difficulty to human time horizons, plus Anthropic–Pentagon contract tensions, Anthropic’s report on distillation attacks (DeepSeek/Moonshot/Minimax), and OpenAI’s report on disrupting malicious use.

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:
(00:00:10) Intro / Banter
(00:01:52) News Preview
Tools & Apps
(00:03:20) Anthropic releases Sonnet 4.6 | TechCrunch
(00:11:24) Google Rolls Out Latest AI Model, Gemini 3.1 Pro - CNET
(00:14:54) Elon Musk says Grok 4.20 public beta is now available: Capabilities of AI chatbot offered by xAI - The Times of India
(00:18:06) Anthropic just released a mobile version of Claude Code called Remote Control | VentureBeat
(00:21:01) Perplexity announces "Computer," an AI agent that assigns work to other AI agents - Ars Technica
Applications & Business
(00:23:40) Meta strikes up to $100B AMD chip deal as it chases 'personal superintelligence' | TechCrunch
(00:27:05) Nvidia challenger AI chip startup MatX raised $500M | TechCrunch
(00:31:00) World Labs lands $1B, with $200M from Autodesk, to bring world models into 3D workflows | TechCrunch
(00:33:07) Simile Raises $100 Million for AI Aiming to Predict Human Behavior
(00:33:52) Stargate AI data centers for OpenAI reportedly delayed by squabbles between partners — sources say OpenAI, Oracle, and SoftBank disagreed on who would have ultimate control of the planned data centers
(00:36:43) China to increase leading-edge chip output by 5x in two years, report claims — aims to lift 7nm and 5nm production to 100,000 wafers per month, targeting half a million monthly by 2030
Research & Advancements
(00:40:33) On Surprising Effectiveness of Masking Updates in Adaptive Optimizers
(00:48:03) Think Deep, Not Just Long: Measuring LLM Reasoning Effort via Deep-Thinking Tokens
(00:54:52) models have some pretty funny attractor states
(01:01:41) When Models Manipulate Manifolds: The Geometry of a Counting Task
(01:05:16) BRIDGE: Predicting Human Task Completion Time From Model Performance
(01:12:00) NESSiE: The Necessary Safety Benchmark -- Identifying Errors that should not Exist
(01:13:15) The least understood driver of AI progress
(01:21:45) The Persona Selection Model: Why AI Assistants might Behave like Humans
Policy & Safety
(01:25:04) Anthropic CEO Amodei says Pentagon's threats 'do not change our position' on AI
(01:33:04) Musk's xAI, Pentagon reach deal to use Grok in classified systems
(01:34:17) Detecting and preventing distillation attacks
(01:38:36) OpenAI details expanding efforts to disrupt malicious use of AI in new report - SiliconANGLE
See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.