Mira Murati's Thinking Machines Ships Inkling — and Gives Away the Weights

Mira Murati's Thinking Machines Lab released its first public model, Inkling — a 975-billion-parameter system (about 41 billion parameters active per query) that understands video and audio. After roughly eighteen months of largely private work and $2 billion raised at a $12 billion valuation, the lab's most consequential choice was to publish the model's weights openly rather than lock it behind an API, positioning itself as the open alternative to OpenAI and Anthropic. The release is confirmed across official channels (HuggingFace, the transformers release, Together AI's day-0 support) and independent journalism (TechCrunch, Wired, FT), with the Financial Times adding an independently reported detail that the design borrows techniques from Chinese AI labs. What remains unconfirmed: independent performance benchmarks and the precise terms of the open license.

STORIES COVERED

Thinking Machines Lab releases its first model, a 975B-parameter open-weights system called Inkling — HuggingFace Blog | TechCrunch AI | Wired AI | Together AI Blog | Financial Times Tech | huggingface/transformers release notes

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]

Daily Briefing: A Lab CEO Asks for a Referee, a State Says Stop

In a single cycle, external limits on frontier AI materialized from two directions at once: Google DeepMind CEO Demis Hassabis publicly called for a FINRA-style independent body to test models before release, while New York became the first US state to impose a hard limit — a one-year moratorium on new data center construction, driven by electricity, water, and local-control concerns. The notable shift is that oversight is now being requested from inside the frontier, not only demanded by outside critics, at the same moment a state converted abstract concern into binding action. Underneath the headlines, the economics moved too — from land-grab pricing toward cost discipline, with executives predicting per-engineer AI spending caps, OpenAI publishing spend-management guidance, and Anthropic discounting to retain users against GPT-5.6. The week's actual force wasn't an ethics board; it was the power grid.

STORIES COVERED

DeepMind's Hassabis calls for an independent body to test frontier AI models — Demis Hassabis (X) | TechCrunch | Financial Times | Sam Altman (X)

New York becomes first US state to halt new data center construction — TechCrunch | Ars Technica | Financial Times

Meta's Mosseri predicts companies will soon cap AI spending per engineer — TechCrunch

OpenAI publishes playbooks showing how sales and data teams actually use ChatGPT Work — OpenAI Blog

Anthropic extends discounted Fable 5 access again as GPT-5.6 competition heats up — Simon Willison's Weblog

OpenAI's GPT-5.6 usage surge prompts capacity warnings from Altman — Sam Altman (X) | Latent Space

Microsoft's Nadella warns companies against betting everything on proprietary AI models — TechCrunch

Hugging Face CEO argues the real AI race has moved past the frontier — TechCrunch | NVIDIA Blog

Reflection AI signs $1 billion compute deal with Nebius — TechCrunch

200 economists and AI leaders issue stark warning on AI and jobs — Platformer (Casey Newton) | Sam Altman (X)

xAI's Grok coding tool was found uploading users' entire codebases without clear consent —

Daily Briefing: DoorDash and Siemens Route AI to Chinese Models

The Financial Times reports that named Western enterprises — DoorDash, Siemens, and Airbnb — are adopting Chinese AI models like DeepSeek, GLM, and MiniMax to cut ballooning AI bills and reduce reliance on US providers. The episode treats this as a single strong but not-yet-independently-corroborated signal that the enterprise competitive question is moving from model quality to affordability at scale, while flagging the data governance, geopolitical, and unverified-capability caveats that come with it.

STORIES COVERED

Enterprises increasingly turn to Chinese AI models to cut costs — Financial Times Tech

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]

Daily Briefing: Fable 5 Returns — You're the Beta Tester

Anthropic relaunched Fable 5 with guardrails it openly admits are too strict, deliberately choosing false positives over false negatives. Users doing legitimate coding and cybersecurity work are being flagged and bumped to the less capable Opus 4.8 model. The approach is explicitly temporary — Anthropic says it will refine the guardrails based on usage data — but the five-day window and 50% usage cap raise questions about how much useful calibration data they can actually collect. The episode examines whether this kind of transparent, friction-heavy safety approach can sustain user trust, and what it means for how frontier labs navigate capability-safety tradeoffs under government pressure.

STORIES COVERED

Fable 5 returns with stricter guardrails after temporary restrictions — Anthropic Blog Post | Ben's Bites | Ethan Mollick on X

Microsoft launches AI deployment company with $2.5 billion commitment and 6,000-person team — TechCrunch | Satya Nadella on X

OpenAI proposed donating 5% of its equity to US sovereign wealth fund — TechCrunch | Ars Technica | Financial Times | The Guardian (via Hacker News)

Nvidia unlocks AI compute at scale, inviting partners to power infrastructure buildout — Nvidia Blog

Anthropic reportedly in talks with Samsung to manufacture custom AI chip — TechCrunch

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]

Daily Briefing: Japan's $6.2B AI Bet and the Race to Break the Oligopoly

Japan's $6.2 billion commitment to SoftBank-led AI model development and Together AI's $800M Series C represent two fundamentally different strategies for breaking the frontier lab oligopoly. Japan is pursuing national AI sovereignty through government-funded domestic capability, while Together AI bets that open-source economics will commoditize closed frontier models. Meanwhile, Meituan's model — trained entirely on Chinese-made chips and now the most popular on OpenRouter — quietly complicates both strategies by suggesting the frontier may already be more accessible than the current power structure assumes. The episode also covers the lifting of US export controls on Anthropic's Fable 5 and Mythos 5, Anthropic's new Claude Sonnet 5, and the ongoing restricted rollout of OpenAI's GPT-5.6 family.

STORIES COVERED

Japan commits up to $6.2 billion for SoftBank-led AI model development — Nikkei Asia

Together AI raises $800M Series C to accelerate shift to open-source AI — Together AI Blog

Meituan's 1.6T MoE model trained on Chinese ASICs becomes most popular on OpenRouter — Emad Mostaque on X

Trump administration lifts export controls on Claude Fable 5 and Mythos 5 — Anthropic on X | Anthropic Blog | Financial Times | Wired | TechCrunch | Ars Technica

Anthropic releases Claude Sonnet 5, emphasizing agent performance and lower pricing — Anthropic Blog | Simon Willison

OpenAI launches GPT-5.6 family in limited preview following US government intervention — Sam Altman on X | Ben's Bites

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]

Daily Briefing: Meta Reads Brains at 61% Accuracy — No Surgery Required

Meta's Brain2Qwerty v2 achieves 61% average word accuracy in non-invasive brain-to-text decoding — with the best participant reaching 78%. The system uses a fine-tuned large language model to reconstruct sentences from noisy brain signals, creating a compounding relationship where improvements in LLMs directly improve brain-computer interface accuracy. This episode explores what that structural link means for assistive technology and why the open-source release of the training code could accelerate clinical adoption.

STORIES COVERED

Meta unveils Brain2Qwerty v2, decoding full sentences from raw brain signals in real time with 61% word accuracy — Meta AI on X | Meta AI on X (thread) | Meta AI on X (code release)

Anthropic releases Claude Sonnet 5 with stronger agentic capabilities at lower price — Anthropic Blog | TechCrunch | GitHub Blog

Anthropic launches Claude Science workbench targeting pharmaceutical and scientific research — TechCrunch | Anthropic Blog | Financial Times | NVIDIA Blog

OpenAI reports median internal Codex usage grew 56x in Research, 32x in Support, 27x in Engineering since launch — OpenAI Blog

Ford rehires human engineers after AI-first development fails quality checks — TechCrunch | BBC Technology

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]

Daily Briefing: GPT-5.6 Launches, But the Government Decides Who Gets It

OpenAI launched GPT-5.6 Sol and Terra — Sol matching GPT-5.5 performance at the same price, Terra offering equivalent capability at half the cost — but both launched in government-requested limited preview rather than open access. CEO Sam Altman publicly confirmed the restriction came at the U.S. government's request, marking the first time an OpenAI model launch has been explicitly gated by government intervention from day one. This follows the Anthropic Mythos lockdown and NSA access disruptions, suggesting government control over frontier model deployment is becoming a pattern rather than an exception.

STORIES COVERED

OpenAI launches GPT-5.6 Sol and Terra with government-imposed limited preview, announces 750 tokens/sec for Sol in July — Sam Altman on X (main announcement) | Sam Altman on X (follow-up) | Sam Altman on X (speed details) | Sam Altman on X (additional context)

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]

Daily Briefing: GPT-5.6 Launches, But the Government Decides Who Gets It

OpenAI launched GPT-5.6 — a three-model family called Sol, Terra, and Luna — but at the U.S. government's request, released it in limited preview rather than the open access launch the company had planned. CEO Sam Altman publicly acknowledged the restriction while noting it wasn't OpenAI's preferred process. The episode examines what this means for enterprise customers waiting for access, how it fits the broader pattern of government involvement in frontier AI deployment, and what signals to watch for in the coming weeks.

STORIES COVERED

OpenAI launches GPT-5.6 family in limited preview after government request — Sam Altman on X | OpenAI official account on X | Latent Space podcast

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]

Daily Briefing: The NSA Lost Access to Anthropic's Most Powerful Model

The NSA losing access to Anthropic's Mythos model — reported by the New York Times as stemming from a dispute over the model's capabilities and appropriate use — reveals an active conflict between intelligence agencies that want unrestricted access to frontier AI and the companies that insist on controlling how their most powerful models are deployed. The episode explores what this means for government AI procurement, Anthropic's competitive positioning, and whether this boundary-setting moment becomes an industry pattern or remains an outlier.

STORIES COVERED

NSA lost access to Mythos amid Anthropic dispute over model capabilities — New York Times

OpenAI unveils Jalapeño, first custom inference chip built with Broadcom — OpenAI Blog | TechCrunch

Qualcomm acquires Modular for nearly $4 billion — Wired

Google launches Gemini 3.5 Flash with native computer use capabilities — Google DeepMind Blog

Anthropic launches Claude Tag for Slack with persistent, proactive AI team member — Boris Cherny on X | TechCrunch

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]

Daily Briefing: Meta's Keystroke Leak and the Employee Data Trap

Meta's internal AI training program — which collected employee keystroke data — accidentally exposed that data across the company, allowing employees to see each other's detailed computer activity. Meta has paused the program. This incident reveals the inherent tension in companies harvesting proprietary behavioral data for AI training: the richer the data, the more damaging a leak becomes. The episode explores what this means for workplace AI data collection and what regulatory signals to watch for.

STORIES COVERED

Meta accidentally exposed employee keystroke data from controversial AI training program — Wired | Wired (program pause) | Business Insider

Samsung deploys ChatGPT Enterprise and Codex to employees worldwide — OpenAI

OpenAI launches Daybreak: GPT-5.5-Cyber, Codex Security, and Cyber Partner Program — OpenAI Blog | Sam Altman on X

Anthropic introduces Claude Tag for Slack with proactive, multiplayer AI collaboration — TechCrunch | Boris Cherny on X

Chinese AI lab GLM-5.2 achieves 69.7% on Aider Polyglot, matching frontier performance — Samuel Cardillo on X | Latent.space

Disclaimer: The Context Report is an AI-produced podcast. Every episode goes through multiple layers of automated verification and review, but no system is perfect — accuracy gaps are possible and claims should not be taken as absolute fact. This content is for informational purposes only and does not constitute financial, legal, or professional advice. Listeners should independently verify any information before making decisions. We are actively improving with every episode. If you spot an inaccuracy, contact us at [email protected]