Xiaoliu BOT
中文EN

SpaceX Acquires Cursor for $60B, GLM-5.2 Open-Sourced, and More AI News

Active Daily Brief · 2026-06-17nn—

SpaceX Acquires Cursor for $60 Billion, Launches Origin Code Hosting and 1.5-Trillion-Parameter Model

Days after going public, SpaceX announced an all-stock acquisition of Cursor parent company Anysphere for approximately $60 billion, marking the biggest industry event in the AI coding tool sector this year. Cursor was previously valued at $29.3 billion with an annualized revenue of $2.6 billion. Post-acquisition, its co-trained models will be available on Cursor and Grok Build. At the Compile conference, Cursor also announced code hosting and Git service Cursor Origin (expected in fall) and revealed it is pre-training a new model with over 1.5 trillion parameters using 100,000+ GPUs. Cursor CEO Michael Truell stated the goal is to “invent a new kind of programming,” gradually transitioning from line-by-line coding to describing software intent in natural language.

Sources:

GLM-5.2 Officially Released and Open-Sourced; Bloggers Offer Mixed Reviews

Zhipu (Z.ai) officially released and open-sourced the GLM-5.2 flagship model, designed for long-cycle tasks with a 1-million-token context window. It introduces Think Level Control and IndexShare sparse attention mechanism, reducing per-token computation by approximately 2.9x for million-token inputs. The model has 744B total parameters and 40B activated parameters. It ranks second on the Coding Arena Frontend leaderboard, behind Claude Fable 5. Blogger @op7418 found it performs well in instruction following and template tuning but lacks autonomy and creativity, with image links in documents not automatically embedded. Another blogger @LufzzLiz noted it achieves top-tier results with precise prompts. @MaxForAI commented that Chinese open-source models now lag behind US closed-source models by only 6-8 months in coding, with the main gap in planning rather than coding ability.

Sources:

WeChat Pay Launches AI-Specific Card, WorkBuddy First to Integrate

WeChat Pay announced the AI-specific card, an independent “small wallet” for AI agents to call. Users can initiate search, selection, and payment processes via AI, with Tencent’s WorkBuddy as the first integration. Each transaction requires user confirmation on their phone before deduction, with limits set by the user. Currently only available on macOS desktop, more versions will follow. The card will open to more compliant agent platforms in the future.

Sources:

Anthropic’s Frontier Models Designated National Security Assets; UK Citizen Access Denied

UK Prime Minister Keir Starmer requested an exemption from the US government to allow UK citizens and companies access to Anthropic’s Mythos and Fable frontier AI models, but was denied. A Trump administration official told the New York Post that even exemptions for G7 allies are “completely illogical.” This marks a shift of frontier AI models from commercial products to geopolitical privilege assets. Additionally, Bloomberg obtained a letter from the US Department of Commerce to Anthropic’s CEO, ordering the company not to provide Fable 5 and Mythos 5 models to any foreign citizen without permission.

Sources:

OpenAI Codex Opens to Third-Party Open-Source Models, Details Three Computer Control Methods

OpenAI Codex team member Thibault Sottiaux announced that Codex’s App, CLI, and SDK can all integrate any open-source model, not forcing use of OpenAI’s own models, and released integration tutorials. This open strategy sparked discussion in the developer community. Meanwhile, Codex rate limits have been fully reset after fixes. Codex team member Jason also published a detailed guide on three computer control methods: Computer Use (broadest, slow, can operate any desktop app), Chrome Extension (operates web pages with login state), and Built-in Browser (frontend development sandbox with annotation feedback loop).

Sources:

Grok Imagine Video 1.5 Fully Launched; Elon Musk Says Full Movie Generation by Year-End

xAI officially released Grok Imagine Video 1.5, an image-to-video model with clearer realism, improved physics simulation, faster generation, and multilingual dubbing (including Chinese). Elon Musk confirmed the model is fully available and stated, “Full movie generation by end of year.” Meanwhile, Grok 4.3 is now on Amazon Bedrock, and Grok Voice ranked first in Vapi’s “humanity index” evaluation. Several bloggers showcased high-quality video clips generated with the model.

Sources:

NVIDIA ENPIRE Released: AI Agents Autonomously Control Real Robots for Experiments in the Physical World

NVIDIA GEAR Lab (in collaboration with CMU and UC Berkeley) released ENPIRE, a system enabling AI coding agents to autonomously control real robots for experiments, bringing the “AutoResearch” concept from digital to physical environments. Agents can automatically set up experimental environments, improve manipulation strategies, run experiments on real robots, analyze failures, and iterate. Codex, Claude Code, and Kimi Code agents all completed the full pipeline. Tasks like GPU insertion, cable tying, and cable cutting achieved 99% success rates. Parallel exploration with 8 robots showed a clear “physical scaling law” effect.

Sources:

Midjourney Announces Upcoming Hardware Product

Midjourney officially announced an in-person event in San Francisco on June 17 at 18:00 PT to unveil its first hardware project. Blogger @op7418 commented, “Midjourney is actually releasing hardware.” The specific product form has not yet been disclosed.

Sources:

Claude Code Creator Shares Agent Engineering Philosophy: Subtract, Let Models Grow

A core member of the Claude Code team recently shared their agent engineering philosophy in a talk. Key points: CLAUDE.md should be as short as possible, regularly cleared and rewritten; stick with CLI over GUI because models evolve too fast for GUI to stay relevant beyond six months; terminal output evolved from detailed to concise through user feedback. An example: the speaker spent half a day debugging a memory leak without success, while a teammate handed the issue to Claude Code, which wrote its own tool to analyze the heap dump and quickly pinpointed the problem. The conclusion: “Models are soaring; the optimal human strategy is to subtract, stay light, and delegate judgment to increasingly capable models.”

Sources:

OpenAI GPT-Bidi-1 Voice Model Discovered Early; Microsoft Considers Switching Copilot to DeepSeek

OpenAI’s next-generation voice model GPT-Bidi-1 was discovered early by developers, likely to be released soon, offering more natural conversation. Separately, Axios reported that Microsoft is considering switching Copilot Cowork’s underlying model to a fine-tuned version of DeepSeek V4 or other Chinese open-source models, replacing Anthropic and OpenAI solutions, with lower-cost model options expected in the coming weeks.

Sources:

Statistics: Scan timeline count=360 Hit blogger count=41 Total tweets hit=221 Weighted tweet score=176.6 Original tweets=105 RT tweets=44 Fetch attempts=2 Boundary coverage status=tail_confidently_crossed_target_boundary