# forck.live — 5–11 August 2024 > 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 2024-W32, 5–11 August 2024, 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/2024-W32 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### Vercel — Update regarding Vercel service disruption on August 7, 2024 - Company: Vercel (vercel.com) - Announced: 2024-08-09T13:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://vercel.com/blog/update-regarding-vercel-service-disruption-on-august-7-2024 - Record: https://forck.live/items/20959-update-regarding-vercel-service-disruption-on-august-7-2024 - Subject: v0 / AI SDK / AI Gateway Vercel reported a 26-minute service disruption on August 7, 2024, caused by an upstream compute provider erroneously blocking traffic. The outage affected 61% of Edge Middleware and Edge Function invocations, while Serverless Functions and static files remained unaffected. Vercel is accelerating redundancy and infrastructure unification efforts to prevent recurrence. ### Alibaba — Qwen2-Audio: Chat with Your Voice! - Company: Alibaba (alibaba.com) - Announced: 2024-08-09T08:18:19+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://qwenlm.github.io/blog/qwen2-audio/ - Record: https://forck.live/items/2310-qwen2-audio-chat-with-your-voice - Subject: Qwen Alibaba releases Qwen2-Audio, the next version of Qwen-Audio, capable of accepting audio and text inputs and generating text outputs. ### Replicate — Replicate Intelligence #10 - Company: Replicate (replicate.com) - Announced: 2024-08-09 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://replicate.com/blog/replicate-intelligence-2024-08-09 - Record: https://forck.live/items/2519-replicate-intelligence-10 - Subject: Platform Replicate introduced image-to-image generation support for the FLUX.1 [dev] model, allowing users to start with an image and a prompt to control style and composition. ### OpenAI — Zico Kolter Joins OpenAI’s Board of Directors - Company: OpenAI (openai.com) - Announced: 2024-08-08T12:00:00+00:00 - Category: funding-company-news - Coverage: not counted - Announcement: no - Group: routine - Source: https://openai.com/index/zico-kolter-joins-openais-board-of-directors - Record: https://forck.live/items/726-zico-kolter-joins-openai-s-board-of-directors - Subject: GPT / ChatGPT / API Zico Kolter joins OpenAI's board of directors, bringing expertise in AI safety and alignment, and will also join the Safety & Security Committee. ### Hugging Face — XetHub is joining Hugging Face! - Company: Hugging Face (huggingface.co) - Announced: 2024-08-08 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/xethub-joins-hf - Record: https://forck.live/items/1842-xethub-is-joining-hugging-face - Subject: Platform Hugging Face acquired XetHub, a Seattle-based company focused on scaling Git for large AI datasets and models. ### Alibaba — Introducing Qwen2-Math - Company: Alibaba (alibaba.com) - Announced: 2024-08-07T16:00:00+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://qwenlm.github.io/blog/qwen2-math/ - Record: https://forck.live/items/2311-introducing-qwen2-math - Subject: Qwen Alibaba introduces math-specific large language models from the Qwen2 series: Qwen2-Math and Qwen2-Math-Instruct-1, focusing on arithmetic and mathematical reasoning. The models primarily support English, with bilingual (English and Chinese) versions to be released soon. ### OpenAI — Pairing data with APIs to unlock customer value - Company: OpenAI (openai.com) - Announced: 2024-08-07T16:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://openai.com/index/rakuten-2024 - Record: https://forck.live/items/728-pairing-data-with-apis-to-unlock-customer-value - Subject: GPT / ChatGPT / API Rakuten is using AI to pair data and unlock customer insights and value. ### LG AI Research — Toward Expert AI : EXAONE 3.0 and ChatEXAONE - Company: LG AI Research (lgresearch.ai) - Announced: 2024-08-07 - Category: open-weight-release - Coverage: not counted - Announcement: yes - Group: models - Source: https://www.lgresearch.ai/news/view?seq=462 - Record: https://forck.live/items/11022-toward-expert-ai-exaone-3-0-and-chatexaone - Subject: EXAONE LG AI Research released EXAONE 3.0, a lineup of models of various sizes, and open-sourced the 7.8B model for research use. The company also unveiled ChatEXAONE, a closed enterprise AI agent built on EXAONE 3.0, initially available to LG Group employees. The 7.8B model is described as the most versatile in the lineup, balancing performance and affordability. ### LG AI Research — Introducing ChatEXAONE : Enterprise AI Agent Boosting Corporate Efficiency with Expert AI Insights - Company: LG AI Research (lgresearch.ai) - Announced: 2024-08-07 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.lgresearch.ai/blog/view?seq=458 - Record: https://forck.live/items/4840-introducing-chatexaone-enterprise-ai-agent-boosting-corporate-efficiency-with - Subject: EXAONE LG AI Research announces ChatEXAONE, an enterprise AI agent built on EXAONE 3.0, designed to help enterprise users work more efficiently and provide expert-level insights. It features professional personalization, document and web-based QA, multimodality (speech and image understanding), and expert modules (coding assistance, data analysis and visualization). The agent is initially provided to LG Group employees. ### LG AI Research — EXAONE 3.0 : Showcasing Our First Open-Source LLM with Global Top-Level Performance - Company: LG AI Research (lgresearch.ai) - Announced: 2024-08-07 - Category: open-weight-release - Coverage: not counted - Announcement: yes - Group: models - Source: https://www.lgresearch.ai/blog/view?seq=460 - Record: https://forck.live/items/4839-exaone-3-0-showcasing-our-first-open-source-llm-with-global-top-level - Subject: EXAONE LG AI Research announces the release of EXAONE 3.0, specifically the 7.8B Instruction Tuned language model as an open-source model for research. The model achieves top global performance in English and Korean benchmarks, demonstrates significant cost reduction (to 6% of the cost of EXAONE 1.0), and includes ethical transparency through red teaming and disclosure of improvement areas. ### LG AI Research — Toward Expert AI : EXAONE 3.0 and ChatEXAONE - Company: LG AI Research (lgresearch.ai) - Announced: 2024-08-07 - Category: open-weight-release - Coverage: not counted - Announcement: yes - Group: models - Source: https://www.lgresearch.ai/blog/view?seq=462 - Record: https://forck.live/items/4838-toward-expert-ai-exaone-3-0-and-chatexaone - Subject: EXAONE LG AI Research introduces EXAONE 3.0, a new model that outperforms similarly sized open models in benchmarks. They open-source the EXAONE 3.0 7.8B model for research purposes. They also unveil ChatEXAONE, an enterprise AI agent built on EXAONE 3.0, for improving productivity of enterprise employees, initially available to LG Group users. ### Mistral AI — Build, tweak, repeat - Company: Mistral AI (mistral.ai) - Announced: 2024-08-07 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://mistral.ai/news/build-tweak-repeat - Record: https://forck.live/items/3343-build-tweak-repeat - Subject: Mistral / Le Chat Mistral AI announced a research post titled 'Build, tweak, repeat' on August 7, 2024. ### Vercel — How to integrate AI into your business - Company: Vercel (vercel.com) - Announced: 2024-08-06T13:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://vercel.com/blog/how-to-integrate-ai-into-your-business - Record: https://forck.live/items/20961-how-to-integrate-ai-into-your-business - Subject: v0 / AI SDK / AI Gateway Vercel published a guide and framework for identifying and evaluating AI use cases in business, including the Vercel Impact Assessment for AI Initiatives (VIA-AI) which scores projects on value, feasibility, alignment, and other criteria. The guide also highlights how Vercel's Frontend Cloud and AI SDK can accelerate AI feature development. ### Vercel — Vercel AI SDK 3.3 - Company: Vercel (vercel.com) - Announced: 2024-08-06T13:00:00+00:00 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/vercel-ai-sdk-3-3 - Record: https://forck.live/items/20960-vercel-ai-sdk-3-3 - Subject: v0 / AI SDK / AI Gateway Vercel AI SDK 3.3 adds experimental tracing via OpenTelemetry, multi-modal file attachments for useChat, a useObject hook for streaming structured JSON to clients, and additional LLM settings. It also introduces AWS Bedrock and Chrome AI model providers. ### OpenAI — Introducing Structured Outputs in the API - Company: OpenAI (openai.com) - Announced: 2024-08-06T10:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/introducing-structured-outputs-in-the-api - Record: https://forck.live/items/729-introducing-structured-outputs-in-the-api - Subject: GPT / ChatGPT / API OpenAI introduces Structured Outputs in the API, ensuring model outputs reliably adhere to developer-supplied JSON Schemas. ### Hugging Face — Introducing TextImage Augmentation for Document Images - Company: Hugging Face (huggingface.co) - Announced: 2024-08-06 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/doc_aug_hf_alb - Record: https://forck.live/items/1844-introducing-textimage-augmentation-for-document-images - Subject: Platform Hugging Face introduces a multimodal text-image augmentation pipeline for document images, developed in collaboration with Albumentations AI. The pipeline modifies both image content and text annotations simultaneously, supporting augmentations like random insertion, deletion, swap, and stopword replacement to improve VLM fine-tuning on limited datasets. ### Hugging Face — 2024 Security Feature Highlights - Company: Hugging Face (huggingface.co) - Announced: 2024-08-06 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/2024-security-features - Record: https://forck.live/items/1843-2024-security-feature-highlights - Subject: Platform Security features available on the Hub as of August 6, 2024, including fine-grained tokens, two-factor authentication, commit signing, organizational access controls, automated security scanning, and enterprise features like single sign-on.