# forck.live — 31 March – 6 April 2025

> AI lab announcements, news reports and podcasts.

forck.live follows what companies announce about their own products. Press may
corroborate an official story; it never leads and never appears as an item of its
own. A missing date stays missing. A title is the publisher's words — where a
company does not publish in English, an English rendering is listed separately
as `Title in English` and named as this site's translation, never as theirs.

This file is one issue of the weekly archive: every confirmed
first-party announcement of ISO week 2025-W14, 31 March – 6 April 2025, newest
first. The window never moves, so this URL always names the same seven days —
but a post published inside them and detected later is added when it arrives,
so a closed week can still gain entries.

Every entry is one first-party announcement, and carries the same eight fields
in this order before any others:

    Company · Announced · Category · Coverage · Announcement · Group · Source · Record

`Source` is the company's own page the announcement was traced to and `Record`
is this site's page for it. Both are safe to cite.

`Coverage` is a number of outlets, or `not counted` — which means this record
reached no count of its own, never that nobody wrote about it.

`Category` is this tracker's own analysis, and is authoritative wherever it is
stated. `not stated` means either that nothing has analysed the post yet or
that the analysis placed it outside this taxonomy; neither is a verdict about
the announcement.

`Announcement` is whether the post is news of something the company did, rather
than a how-to, a customer story, a changelog listing or a staffing note.

`Group` is which of this record's groups the entry falls in, carried on the
entry itself so it holds where the groups are headings and where they are not.
One of: `models`, `covered`, `announcements`, `routine`.

Anything after `Record` is held detail that only some announcements carry.

- This week's page: https://forck.live/week/2025-W14
- The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md
- Catalogue: https://forck.live/llms.txt

## Announcements

### Hugging Face — Welcome Llama 4 Maverick & Scout on Hugging Face

- Company: Hugging Face (huggingface.co)
- Announced: 2025-04-05
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://huggingface.co/blog/llama4-release
- Record: https://forck.live/items/1718-welcome-llama-4-maverick-scout-on-hugging-face
- Subject: Platform

Hugging Face announces the availability of Meta's Llama 4 Maverick and Llama 4 Scout models on its platform, with details on architecture, context lengths, and integrations.

### Supabase — Top 10 Launches of Launch Week 14

- Company: Supabase (supabase.com)
- Announced: 2025-04-04T07:00:02+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://supabase.com/blog/launch-week-14-top-10
- Record: https://forck.live/items/21360-top-10-launches-of-launch-week-14
- Subject: Supabase

Supabase announced ten product updates during Launch Week 14, including the ability to deploy Edge Functions from the Dashboard, extended Realtime Broadcast from database triggers, routing Data API requests to the nearest Read Replica, an official UI Library, an MCP server for AI tools, declarative schemas, Clerk Auth integration, a Postgres Language Server, dedicated connection poolers, and a partnership invitation for AI builders.

### Supabase — Supabase MCP Server

- Company: Supabase (supabase.com)
- Announced: 2025-04-04T07:00:01+00:00
- Category: developer-tool-release
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/mcp-server
- Record: https://forck.live/items/21361-supabase-mcp-server
- Subject: Supabase

Supabase launched an official MCP (Model Context Protocol) server that allows AI tools like Cursor and Claude to connect directly with Supabase. The server provides over 20 tools for tasks such as launching databases, managing tables, fetching configuration, and querying data. Setup requires a personal access token for authentication.

### Supabase — Data API Routes to Nearest Read Replica

- Company: Supabase (supabase.com)
- Announced: 2025-04-04
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/data-api-nearest-read-replica
- Record: https://forck.live/items/21362-data-api-routes-to-nearest-read-replica
- Subject: Supabase

Supabase released geo-routing for Data API requests, automatically directing read-only PostgREST requests to the geographically closest read replica to reduce latency. The feature is available by default when using the API load balancer endpoint and requires no configuration changes beyond updating the project URL.

### Genspark — Why I Killed Our AI Search Product With 5 Million Users?

- Company: Genspark (genspark.ai)
- Announced: 2025-04-04
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://www.genspark.ai/blog/genspark-kill-aisearch
- Record: https://forck.live/items/18291-why-i-killed-our-ai-search-product-with-5-million-users
- Subject: Genspark

product Kay, Co-founder and CTO of Genspark

### LG AI Research — Understanding Creative Connections Through Relational Artifacts

- Company: LG AI Research (lgresearch.ai)
- Announced: 2025-04-04
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.lgresearch.ai/news/view?seq=546
- Record: https://forck.live/items/11008-understanding-creative-connections-through-relational-artifacts
- Subject: EXAONE

LG AI Research and Parsons School of Design conducted a study on human-AI creative collaboration using the AI agent ChatEXAONE. The experiment paired human participants with an AI partner whose responses were read aloud by a human proxy, and found that while participants sometimes felt a sense of connection, the AI lacked genuine emotional engagement and natural conversational flow.

### LG AI Research — Understanding Creative Connections Through Relational Artifacts

- Company: LG AI Research (lgresearch.ai)
- Announced: 2025-04-04
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.lgresearch.ai/blog/view?seq=546
- Record: https://forck.live/items/4798-understanding-creative-connections-through-relational-artifacts
- Subject: EXAONE

This research project, conducted by Parsons School of Design and LG AI Research, uses an experimental design where human participants converse with the AI agent ChatEXAONE through a human proxy to explore whether AI can foster creative intimacy. The study examines how AI responses shape human expression and redefine agency, finding that while AI can serve as a catalyst for reflection, it lacks genuine emotional depth.

### Hugging Face — Journey to 1 Million Gradio Users!

- Company: Hugging Face (huggingface.co)
- Announced: 2025-04-04
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://huggingface.co/blog/gradio-1m
- Record: https://forck.live/items/1719-journey-to-1-million-gradio-users
- Subject: Platform

Gradio, a Python library for building AI web apps, has reached over 1 million monthly developers.

### Runway — Towards a new media ecosystem with world simulators

- Company: Runway (runway.com)
- Announced: 2025-04-03T23:23:58.290000+00:00
- Category: funding-company-news
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://runway.com/news/runway-series-d-funding
- Record: https://forck.live/items/7912-towards-a-new-media-ecosystem-with-world-simulators
- Subject: Gen / Aleph

Runway announced over $300M in Series D funding led by General Atlantic to expand Runway Studios and advance research in media generation. The company also unveiled Gen-4, its next-generation series of AI models for media generation and world consistency.

### Vercel — How PAIGE grew revenue by 22% with Shopify, Next.js, and Vercel

- Company: Vercel (vercel.com)
- Announced: 2025-04-03T13:00:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://vercel.com/blog/how-paige-grew-revenue-by-22-with-shopify-next-js-and-vercel
- Record: https://forck.live/items/20890-how-paige-grew-revenue-by-22-with-shopify-next-js-and-vercel
- Subject: v0 / AI SDK / AI Gateway

PAIGE, a denim and apparel retailer, re-architected its ecommerce platform using Shopify, Next.js, and Vercel. The new headless stack resulted in a 22% increase in revenue and a 76% increase in conversion rates during Black Friday Cyber Monday, with no downtime.

### Supabase — Declarative Schemas for Simpler Database Management

- Company: Supabase (supabase.com)
- Announced: 2025-04-03
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/declarative-schemas
- Record: https://forck.live/items/21363-declarative-schemas-for-simpler-database-management
- Subject: Supabase

Supabase released declarative schemas, allowing developers to define their database structure in centralized .sql files that can be version-controlled. The feature automatically generates migration files and simplifies schema changes by providing a single place to maintain tables, views, functions, RLS policies, and other database entities. Supabase has used the same internal tooling for the past two years and is now making it available to all users via the Supabase CLI.

### Upstage — Demo now open! Unleash information from any document

- Company: Upstage (upstage.ai)
- Announced: 2025-04-03
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://www.upstage.ai/blog/en/demo-now-open-unleash-information-from-any-document
- Record: https://forck.live/items/18473-demo-now-open-unleash-information-from-any-document
- Subject: Solar Pro / Solar Mini

Upstage released a demo of its upcoming universal information extraction API, which turns unstructured documents into structured data. The demo walks users through a five-step process: uploading a document, auto-generating a schema, editing the schema, confirming extraction results, and testing integration via auto-generated code snippets. The full API is not yet available; early access is by waitlist.

### Upstage — Upstage and Fusic Sign Strategic Partnership

- Company: Upstage (upstage.ai)
- Announced: 2025-04-03
- Category: partnership-acquisition
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.upstage.ai/news/upstage-fusic-partnership
- Record: https://forck.live/items/18435-upstage-and-fusic-sign-strategic-partnership
- Subject: Solar Pro / Solar Mini

Upstage and Fusic have signed a strategic partnership to accelerate generative AI adoption in Japan. Upstage will provide its LLM series and the Syn model, while Fusic contributes its local customer network and regional expertise. The collaboration includes technical training, advisory support, joint PoC initiatives, and co-hosted events.

### Hugging Face — The NLP Course is becoming the LLM Course

- Company: Hugging Face (huggingface.co)
- Announced: 2025-04-03
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://huggingface.co/blog/llm-course
- Record: https://forck.live/items/1720-the-nlp-course-is-becoming-the-llm-course
- Subject: Platform

Hugging Face is renaming and expanding its NLP Course to the LLM Course, adding new chapters on fine-tuning, reasoning models, and more, while maintaining classic NLP content.

### Hugging Face — Efficient Request Queueing – Optimizing LLM Performance

- Company: Hugging Face (huggingface.co)
- Announced: 2025-04-02T13:33:53+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://huggingface.co/blog/tngtech/llm-performance-request-queueing
- Record: https://forck.live/items/1721-efficient-request-queueing-optimizing-llm-performance
- Subject: Platform

The problem of request queuing in LLM inference engines and presents a solution using fair scheduling with separate queues per user and model, and dynamic rate limiting to keep the backend queue short.

### OpenAI — New commission to provide insight as OpenAI builds the world’s best-equipped nonprofit

- Company: OpenAI (openai.com)
- Announced: 2025-04-02T12:00:00+00:00
- Category: funding-company-news
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://openai.com/index/nonprofit-commission-guidance
- Record: https://forck.live/items/590-new-commission-to-provide-insight-as-openai-builds-the-world-s-best-equipped
- Subject: GPT / ChatGPT / API

OpenAI announces a commission to provide guidance as it aims to build the world's best-equipped nonprofit, combining financial resources with AI technology.

### OpenAI — PaperBench: Evaluating AI’s Ability to Replicate AI Research

- Company: OpenAI (openai.com)
- Announced: 2025-04-02T10:15:00+00:00
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://openai.com/index/paperbench
- Record: https://forck.live/items/591-paperbench-evaluating-ai-s-ability-to-replicate-ai-research
- Subject: GPT / ChatGPT / API

OpenAI introduces PaperBench, a benchmark to evaluate AI agents' ability to replicate state-of-the-art AI research.

### Supabase — Realtime: Broadcast from Database

- Company: Supabase (supabase.com)
- Announced: 2025-04-02T07:00:01+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/realtime-broadcast-from-database
- Record: https://forck.live/items/21364-realtime-broadcast-from-database
- Subject: Supabase

Supabase added Broadcast from Database, a new Realtime feature that lets developers send database changes to clients using SQL triggers and functions. It offers lower latency, selective column payloads, and the ability to target specific actions (INSERT, UPDATE, DELETE, TRUNCATE) compared to the existing Postgres Changes polling method.

### OpenAI — Our response to the UK’s copyright consultation

- Company: OpenAI (openai.com)
- Announced: 2025-04-02T07:00:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://openai.com/global-affairs/response-to-uk-copyright-consultation
- Record: https://forck.live/items/592-our-response-to-the-uk-s-copyright-consultation
- Subject: GPT / ChatGPT / API

OpenAI submitted recommendations for pro-innovation policies to help the UK become the AI capital of Europe, in response to the UK's copyright consultation.

### Supabase — Keeping Tabs on What's New in Supabase Studio

- Company: Supabase (supabase.com)
- Announced: 2025-04-02
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/tabs-dashboard-updates
- Record: https://forck.live/items/21365-keeping-tabs-on-what-s-new-in-supabase-studio
- Subject: Supabase

Supabase Studio added tabs to the Table Editor and SQL Editor, customizable reports with resizable and reorderable chart blocks, inline SQL execution within report blocks, an inline SQL editor accessible from anywhere in the dashboard, multiple AI Assistant chats with project-scoped history, and improved log details with stacked charts and HTTP status filters.

### Genspark — Meet Genspark Super Agent，Your Ultimate AI Assistant!

- Company: Genspark (genspark.ai)
- Announced: 2025-04-02
- Category: not stated
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.genspark.ai/blog/genspark-super-agent
- Record: https://forck.live/items/18292-meet-genspark-super-agent-your-ultimate-ai-assistant
- Subject: Genspark

### Krea — Krea Teams: Creative Collaboration Made Easy

- Company: Krea (krea.ai)
- Announced: 2025-04-02
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.krea.ai/blog/introducing-teams
- Record: https://forck.live/items/15096-krea-teams-creative-collaboration-made-easy
- Subject: Krea 2 / AI creative platform

Krea introduced Teams Essentials, a plan that allows users to invite team members, share compute resources, and manage billing from a single owner account.

### Vercel — The no-nonsense guide to composable commerce

- Company: Vercel (vercel.com)
- Announced: 2025-04-01T13:00:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://vercel.com/blog/the-no-nonsense-guide-to-composable-commerce
- Record: https://forck.live/items/20891-the-no-nonsense-guide-to-composable-commerce
- Subject: v0 / AI SDK / AI Gateway

Vercel's guide presents five truths for adopting composable commerce, emphasizing incremental changes focused on business outcomes, retaining monolithic backends when possible, and avoiding unnecessary middleware or redesigns.

### Supabase — Edge Functions: Deploy from the Dashboard + Deno 2.1

- Company: Supabase (supabase.com)
- Announced: 2025-04-01T07:00:01+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/supabase-edge-functions-deploy-dashboard-deno-2-1
- Record: https://forck.live/items/21366-edge-functions-deploy-from-the-dashboard-deno-2-1
- Subject: Supabase

Supabase now allows developers to create, test, edit, and deploy Edge Functions directly from the Supabase Dashboard, removing the previous requirement to install the Supabase CLI and Docker. The Dashboard editor includes syntax highlighting, type-checking, templates for common use cases, and an inline AI Assistant. Additionally, Supabase added Deno 2.1 support for Edge Functions and introduced new APIs for deploying functions without Docker.

### Ant Ling — Ring: A Reasoning MoE LLM Provided and Open-sourced by InclusionAI

- Company: Ant Ling (ant-ling.com)
- Announced: 2025-04-01
- Category: open-weight-release
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://developer.ant-ling.com/en/blogs/ring
- Record: https://forck.live/items/22354-ring-a-reasoning-moe-llm-provided-and-open-sourced-by-inclusionai
- Subject: Ling / Ring / Ming

InclusionAI open-sourced Ring, a reasoning mixture-of-experts large language model derived from Ling, along with several variants including Ring-lite-distill-preview (16.8B total parameters, 2.75B activated). The models are available on Hugging Face and ModelScope under the MIT License.

### Ant Ling — PromptCoT & PromptCoT-Mamba: Advancing the Frontiers of Reasoning

- Company: Ant Ling (ant-ling.com)
- Announced: 2025-04-01
- Category: not stated
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://developer.ant-ling.com/en/blogs/promptcot
- Record: https://forck.live/items/22353-promptcot-promptcot-mamba-advancing-the-frontiers-of-reasoning
- Subject: Ling / Ring / Ming

PromptCoT-Mamba released! Introducing an attention-free foundation model for reasoning tasks.

### Ant Ling — In RL we trust — AReaL v0.2 (Boba) Release

- Company: Ant Ling (ant-ling.com)
- Announced: 2025-04-01
- Category: open-weight-release
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://developer.ant-ling.com/en/blogs/areal-boba
- Record: https://forck.live/items/22345-in-rl-we-trust-areal-v0-2-boba-release
- Subject: Ling / Ring / Ming

Ant Ling released AReaL v0.2 (Boba), an open-source reinforcement learning framework. The release includes a SOTA 7B model for mathematical reasoning (pass@1 61.9 on AIME24, 48.3 on AIME25) and a competitive 32B model trained with only 200 data samples that matches QwQ-32B on AIME 2024. The framework adds SGLang support and achieves 1.5x speed improvement over v0.1 on 7B models.

### Ant Ling — AReaL: Ant Reasoning Reinforcement Learning for LLMs

- Company: Ant Ling (ant-ling.com)
- Announced: 2025-04-01
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://developer.ant-ling.com/en/blogs/areal
- Record: https://forck.live/items/22344-areal-ant-reasoning-reinforcement-learning-for-llms
- Subject: Ling / Ring / Ming

Ant Research released AReaL (Ant Reasoning RL), an open-source fully asynchronous reinforcement learning training system for large reasoning models. The v0.3 release (boba²) achieves a 2.77x speedup over synchronous systems and includes SOTA coding models: a 14B model scoring 69.1 on LCB-v5. The system supports multi-turn agentic RL training and scales from a single node to 1K GPUs.

### Ant Ling — Agentic Learning

- Company: Ant Ling (ant-ling.com)
- Announced: 2025-04-01
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://developer.ant-ling.com/en/blogs/agenticlearning
- Record: https://forck.live/items/22343-agentic-learning
- Subject: Ling / Ring / Ming

Ant Ling introduced AgenticLearning, an open-source agent training paradigm for multi-turn interactions with environments. The project includes two frameworks: RAG-R1 for deep search and reasoning via multi-query parallelism, and FunReason for enhancing function calling capabilities through automated data refinement and a Self-Refinement Multiscale Loss approach. FunReason achieves GPT-4o-comparable performance on BFCL and mitigates catastrophic forgetting on HumanEval and MBPP.

### Supabase — Automatic Embeddings in Postgres

- Company: Supabase (supabase.com)
- Announced: 2025-04-01
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/automatic-embeddings
- Record: https://forck.live/items/21367-automatic-embeddings-in-postgres
- Subject: Supabase

Supabase released automatic embeddings, a feature that automates embedding generation and updates within Postgres using SQL triggers and extensions like pgmq, pg_cron, and pg_net. This eliminates the need for separate external pipelines, reducing drift and complexity. Users can manage embedding jobs entirely in SQL, with support for any external inference API.

### Replicate — Stylized video with Wan2.1

- Company: Replicate (replicate.com)
- Announced: 2025-04-01
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://replicate.com/blog/wan-t2v-loras
- Record: https://forck.live/items/2499-stylized-video-with-wan2-1
- Subject: Platform

Replicate now supports fast inference with LoRAs on the Wan 2.1 model, enabling users to apply premade video styles or train custom ones for stylized video generation.

### Runway — Runway Gen-4: AI Video Generation with World Consistency

- Company: Runway (runway.com)
- Announced: 2025-03-31T16:22:35.195000+00:00
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://runway.com/research/introducing-runway-gen-4
- Record: https://forck.live/items/7913-runway-gen-4-ai-video-generation-with-world-consistency
- Subject: Gen / Aleph

Runway announced Gen-4, a next-generation AI video generation model that can generate consistent characters, objects, and locations across scenes using visual references and instructions without fine-tuning.

### OpenAI — New funding to build towards AGI

- Company: OpenAI (openai.com)
- Announced: 2025-03-31T15:00:00+00:00
- Category: funding-company-news
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://openai.com/index/march-funding-updates
- Record: https://forck.live/items/593-new-funding-to-build-towards-agi
- Subject: GPT / ChatGPT / API

OpenAI announces $40B in new funding at a $300B post-money valuation, aiming to advance AI research, scale compute infrastructure, and deliver more powerful tools for ChatGPT users.

### Supabase — Introducing the Supabase UI Library

- Company: Supabase (supabase.com)
- Announced: 2025-03-31T07:00:01+00:00
- Category: developer-tool-release
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/supabase-ui-library
- Record: https://forck.live/items/21368-introducing-the-supabase-ui-library
- Subject: Supabase

Supabase released an official UI Library of ready-to-use components built on shadcn/ui, compatible with Next.js, React Router, TanStack Start, and plain React. The library includes components for password-based authentication, file uploads, realtime cursor sharing, presence avatars, and chat. Alongside the UI components, Supabase also shipped a curated set of LLM rules for AI code editors tailored to Supabase and Postgres.

### Supabase — Supabase Auth: Bring Your Own Clerk

- Company: Supabase (supabase.com)
- Announced: 2025-03-31
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://supabase.com/blog/clerk-tpa-pricing
- Record: https://forck.live/items/21369-supabase-auth-bring-your-own-clerk
- Subject: Supabase

Supabase expanded its official Third-party Auth integrations to include Clerk, offering a more secure implementation than the previous workaround. The company also reduced pricing for Third-party Auth to match Supabase Auth, with up to 50,000 MAU on the Free plan and 100,000 MAU on the Pro plan, then $0.00325 per MAU above that.

### Isomorphic Labs — Isomorphic Labs announces $600m external investment round

- Company: Isomorphic Labs (isomorphiclabs.com)
- Announced: 2025-03-31
- Category: funding-company-news
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.isomorphiclabs.com/articles/isomorphic-labs-announces-600m-external-investment-round
- Record: https://forck.live/items/19137-isomorphic-labs-announces-600m-external-investment-round
- Subject: Drug Design Engine / AlphaFold

Isomorphic Labs raised $600 million in its first external funding round, led by Thrive Capital with participation from GV and follow-on capital from Alphabet. The company aims to apply its AI drug design engine to accelerate drug discovery across multiple therapeutic areas.

### Cua — Build Your Own Operator on macOS - Part 1

- Company: Cua (cua.ai)
- Announced: 2025-03-31
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://cua.ai/blog/build-your-own-operator-on-macos-1
- Record: https://forck.live/items/16846-build-your-own-operator-on-macos-part-1
- Subject: Cua Driver / Lume / Cua Bench / Cua Fleets

Cua published a tutorial on building a computer-use operator on macOS using OpenAI's computer-use-preview model and the Responses API. The guide covers setting up a macOS virtual machine, connecting the model to it, and implementing a basic interaction loop. The computer-use-preview model is currently limited to OpenAI tier 3+ users and the Responses API.

### Luma — Introducing Camera Motion Concepts

- Company: Luma (lumalabs.ai)
- Announced: 2025-03-31
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://lumalabs.ai/news/camera-motion-concepts
- Record: https://forck.live/items/7989-introducing-camera-motion-concepts
- Subject: Ray / Dream Machine

Luma released Camera Motion Concepts for its Ray2 video generation model, enabling reliable and composable camera motion control. The Concepts can be learned from one or a few examples and combined via natural language at inference time without compromising model quality.

### Hugging Face — How Hugging Face Scaled Secrets Management for AI Infrastructure

- Company: Hugging Face (huggingface.co)
- Announced: 2025-03-31
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://huggingface.co/blog/scaling-secrets-management
- Record: https://forck.live/items/1722-how-hugging-face-scaled-secrets-management-for-ai-infrastructure
- Subject: Platform

Hugging Face published a case study about migrating secrets management to Infisical to improve security and developer workflows.