# forck.live — 20–26 May 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-W21, 20–26 May 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-W21 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### xAI — Series B funding round - Company: xAI (x.ai) - Announced: 2024-05-26 - Category: funding-company-news - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://x.ai/news/series-b - Record: https://forck.live/items/4354-series-b-funding-round - Subject: Grok xAI announces a $6 billion Series B funding round. ### Vercel — Vercel Ship 2024 recap - Company: Vercel (vercel.com) - Announced: 2024-05-24T13:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: no - Group: routine - Source: https://vercel.com/blog/vercel-ship-2024 - Record: https://forck.live/items/20977-vercel-ship-2024-recap - Subject: v0 / AI SDK / AI Gateway Vercel announced new features at Ship 2024, including feature flag integration with Vercel Web Analytics and logs, the Vercel Firewall for traffic management, and improvements to the Vercel Toolbar, Next.js 15 RC, v0, and the AI SDK. ### TII — Falcon 2: An 11B parameter pretrained language model and VLM, trained on over 5000B tokens and 11 languages - Company: TII (tii.ae) - Announced: 2024-05-24T12:00:00+00:00 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://falcon-lm.github.io/blog/falcon2-11b/ - Record: https://forck.live/items/18419-falcon-2-an-11b-parameter-pretrained-language-model-and-vlm-trained-on-over - Subject: Falcon LLM TII released Falcon 2, a new generation of open-source models, starting with an 11B parameter language model (LLM) and an 11B vision-language model (VLM) that can answer queries about images. The LLM was trained on over 5,000 billion tokens of RefinedWeb and curated data, supports 11 languages, and achieves performance comparable to Falcon-40B at a quarter of the size. The VLM integrates a CLIP ViT-L/14 vision encoder with the chat-finetuned LLM. ### Hugging Face — CyberSecEval 2 - A Comprehensive Evaluation Framework for Cybersecurity Risks and Capabilities of Large Language Models - Company: Hugging Face (huggingface.co) - Announced: 2024-05-24 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/leaderboard-llamaguard - Record: https://forck.live/items/1883-cyberseceval-2-a-comprehensive-evaluation-framework-for-cybersecurity-risks - Subject: Platform CyberSecEval 2, a comprehensive evaluation framework for cybersecurity risks and capabilities of Large Language Models (LLMs). It includes benchmarks for insecure code generation, prompt injection, compliance with cyber attack requests, code interpreter abuse, and automated offensive cybersecurity capabilities. The post also provides key insights from evaluating state-of-the-art LLMs, noting a decrease in compliance with cyber attack requests since the first version, but highlighting ongoing challenges with prompt injection and interpreter abuse. ### Hugging Face — Falcon 2: An 11B parameter pretrained language model and VLM, trained on over 5000B tokens and 11 languages - Company: Hugging Face (huggingface.co) - Announced: 2024-05-24 - Category: open-weight-release - Coverage: not counted - Announcement: yes - Group: models - Source: https://huggingface.co/blog/falcon2-11b - Record: https://forck.live/items/1882-falcon-2-an-11b-parameter-pretrained-language-model-and-vlm-trained-on-over - Subject: Platform TII releases Falcon 2, a new generation of open-source models including an 11B parameter LLM and an 11B VLM, trained on over 5000B tokens and supporting 11 languages, with improved performance and multi-modal capabilities. ### Vercel — Introducing new developer tools in the Vercel Toolbar - Company: Vercel (vercel.com) - Announced: 2024-05-23T13:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/introducing-new-developer-tools-in-the-vercel-toolbar - Record: https://forck.live/items/20980-introducing-new-developer-tools-in-the-vercel-toolbar - Subject: v0 / AI SDK / AI Gateway Learn about the new tools for accessibility, interaction timing, and Open Graph images—and the ability to use the toolbar in production. ### Vercel — Shipping safer and smarter: Integrating feature flags deeper in the Vercel workflow - Company: Vercel (vercel.com) - Announced: 2024-05-23T13:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/feature-flags - Record: https://forck.live/items/20979-shipping-safer-and-smarter-integrating-feature-flags-deeper-in-the-vercel - Subject: v0 / AI SDK / AI Gateway Introducing a platform-wide understanding of feature flags and an experimental Next.js design pattern. Feature flags help teams to release with confidence, safely roll out changes, and test efficiently, improving collaboration and accelerating development cycles. ### Vercel — Introducing the Vercel Web Application Firewall - Company: Vercel (vercel.com) - Announced: 2024-05-23T13:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/introducing-the-vercel-waf - Record: https://forck.live/items/20978-introducing-the-vercel-web-application-firewall - Subject: v0 / AI SDK / AI Gateway Vercel introduced a Web Application Firewall (WAF) that provides Layer 7 protection, customizable rules, framework-aware rules, managed rulesets for OWASP Top 10 risks, rate limiting (beta), and observability. The WAF is available on all plans and deploys automatically within the Vercel platform without additional routing or external tools. ### LG AI Research — LG and UNESCO to Create an AI Ethics Platform Together - Company: LG AI Research (lgresearch.ai) - Announced: 2024-05-23 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.lgresearch.ai/news/view?seq=432 - Record: https://forck.live/items/11027-lg-and-unesco-to-create-an-ai-ethics-platform-together - Subject: EXAONE LG AI Research and UNESCO signed an agreement to jointly develop and promote an AI ethics global MOOC. They will form an advisory group and begin curriculum development within the year, aiming for a late 2025 public release. The content will be made available free of charge through UNESCO's Global AI Ethics and Governance Observatory. ### LG AI Research — LG and UNESCO to Create an AI Ethics Platform Together - Company: LG AI Research (lgresearch.ai) - Announced: 2024-05-23 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.lgresearch.ai/blog/view?seq=432 - Record: https://forck.live/items/4853-lg-and-unesco-to-create-an-ai-ethics-platform-together - Subject: EXAONE LG AI Research and UNESCO signed an agreement to jointly develop an AI ethics global MOOC, with an advisory group to begin work within the year, a public release targeted for late 2025, and interim results at the AI Action Summit. The course content will be available free of charge through UNESCO's Global AI Ethics and Governance Observatory. ### Replicate — Shared network vulnerability disclosure - Company: Replicate (replicate.com) - Announced: 2024-05-23 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://replicate.com/blog/shared-network-vulnerability-disclosure - Record: https://forck.live/items/2538-shared-network-vulnerability-disclosure - Subject: Platform Replicate discloses a security vulnerability reported by Wiz that could have allowed malicious models to access sensitive data. The vulnerability was mitigated within 24 hours of speaking with Wiz, and additional mitigations including encrypting all internal traffic and restricting network privileges were deployed. No evidence of exploitation was found. ### Harvey — Mistral AI and Harvey Partnership - Company: Harvey (harvey.ai) - Announced: 2024-05-22T21:31:00+00:00 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.harvey.ai/blog/mistral-announcement - Record: https://forck.live/items/17734-mistral-ai-and-harvey-partnership - Subject: Harvey Harvey announced a partnership with Mistral AI to accelerate support for customers in highly regulated industries, citing Mistral's focus on performance, transparency, and customization. ### OpenAI — A landmark multi-year global partnership with News Corp - Company: OpenAI (openai.com) - Announced: 2024-05-22T13:15:00+00:00 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/news-corp-and-openai-sign-landmark-multi-year-global-partnership - Record: https://forck.live/items/760-a-landmark-multi-year-global-partnership-with-news-corp - Subject: GPT / ChatGPT / API OpenAI and News Corp have entered into a multi-year global partnership to bring premium journalism to OpenAI's generative AI products. ### Upstage — Introducing Solar Mini Chat ja: Expanding language support to Japanese - Company: Upstage (upstage.ai) - Announced: 2024-05-22 - Category: model-update - Coverage: not counted - Announcement: yes - Group: models - Source: https://www.upstage.ai/blog/en/solar-mini-chat-ja - Record: https://forck.live/items/18482-introducing-solar-mini-chat-ja-expanding-language-support-to-japanese - Subject: Solar Pro / Solar Mini Upstage released Solar mini chat ja, a version of its Solar Mini Chat model fine-tuned for Japanese multi-turn conversations. The model supports English, Korean, and Japanese, and is available immediately to existing API users without requiring new keys. ### Hugging Face — Deploy models on AWS Inferentia2 from Hugging Face - Company: Hugging Face (huggingface.co) - Announced: 2024-05-22 - Category: infrastructure-release - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/inferentia-inference-endpoints - Record: https://forck.live/items/1884-deploy-models-on-aws-inferentia2-from-hugging-face - Subject: Platform Hugging Face adds support for AWS Inferentia2 to its Inference Endpoints and SageMaker deployment options, enabling easy deployment of over 100,000 models on Inf2 instances. ### Replit — Using Replit to get our grandfather the daily baseball scores - Company: Replit (replit.com) - Announced: 2024-05-21T16:26:45+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://replit.com/blog/using-replit-to-get-my-grandfather-in-law-the-daily-baseball-scores - Record: https://forck.live/items/17994-using-replit-to-get-our-grandfather-the-daily-baseball-scores - Subject: Replit Agent A Replit user describes building a custom application that emails daily MLB scores to his grandfather-in-law, using Replit Scheduled Deployments, Selenium for screenshots, and Resend for email delivery. The project cost less than $0.50 per month to run. ### OpenAI — OpenAI safety practices - Company: OpenAI (openai.com) - Announced: 2024-05-21T06:00:00+00:00 - Category: safety-policy-update - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/openai-safety-update - Record: https://forck.live/items/761-openai-safety-practices - Subject: GPT / ChatGPT / API OpenAI states that artificial general intelligence must be developed and deployed responsibly. ### Suno — Suno has raised $125 million - Company: Suno (suno.com) - Announced: 2024-05-21 - Category: funding-company-news - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://suno.com/blog/fundraising-announcement-may-2024 - Record: https://forck.live/items/10351-suno-has-raised-125-million - Subject: Music / Studio Suno announced it has raised $125 million in funding to accelerate product development and expand its team. The company, which released its music-generation product eight months prior, reported that 10 million people have already used Suno to create music. ### Hugging Face — From cloud to developers: Hugging Face and Microsoft Deepen Collaboration - Company: Hugging Face (huggingface.co) - Announced: 2024-05-21 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/microsoft-collaboration - Record: https://forck.live/items/1888-from-cloud-to-developers-hugging-face-and-microsoft-deepen-collaboration - Subject: Platform Hugging Face and Microsoft announced a deepened strategic collaboration at Microsoft Build 2024, including expanded Hugging Face Collection in Azure Model Catalog with new models (Llama 3, Mistral 7B, Command R Plus, Qwen 1.5 110B), new AMD MI300X VMs on Azure, open science collaboration with Phi-3 models (MIT license), open source collaboration with ONNX Runtime and WebGPU, and a new Spaces Dev Mode integration with VS Code for developers. ### Hugging Face — Hugging Face on AMD Instinct MI300 GPU - Company: Hugging Face (huggingface.co) - Announced: 2024-05-21 - Category: infrastructure-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/huggingface-amd-mi300 - Record: https://forck.live/items/1887-hugging-face-on-amd-instinct-mi300-gpu - Subject: Platform Hugging Face announces first-class integration of AMD Instinct MI300 GPU servers into the Hugging Face Platform, enabling seamless deployment of AI models on AMD hardware without code changes. The collaboration includes CI/CD testing on MI300 GPUs for Transformers and text-generation-inference, and performance optimizations for inference workloads using techniques like Flash Attention v2, Paged Attention, GPTQ/AWQ compression, and AMD's TunableOp. ### Hugging Face — Build AI on premise with Dell Enterprise Hub - Company: Hugging Face (huggingface.co) - Announced: 2024-05-21 - Category: product-launch - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/dell-enterprise-hub - Record: https://forck.live/items/1886-build-ai-on-premise-with-dell-enterprise-hub - Subject: Platform Hugging Face and Dell announce the Dell Enterprise Hub, a curated experience for training and deploying open models on-premises using Dell platforms, with support for models like Llama 3, Mixtral, and Gemma. ### Hugging Face — Introducing Spaces Dev Mode for a seamless developer experience - Company: Hugging Face (huggingface.co) - Announced: 2024-05-21 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/spaces-dev-mode - Record: https://forck.live/items/1885-introducing-spaces-dev-mode-for-a-seamless-developer-experience - Subject: Platform Hugging Face introduced Spaces Dev Mode, a beta feature for PRO subscribers that allows direct VS Code or SSH connection to a Space, enabling instant code editing without git pushes. ### LG AI Research — 1st Place in EHRSQL 2024: Practical Text-to-SQL System for Electronic Health Records - Company: LG AI Research (lgresearch.ai) - Announced: 2024-05-20 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.lgresearch.ai/blog/view?seq=430 - Record: https://forck.live/items/4854-1st-place-in-ehrsql-2024-practical-text-to-sql-system-for-electronic-health - Subject: EXAONE LG AI Research and KAIST achieved 1st place in the EHRSQL shared task at ClinicalNLP 2024 with a text-to-SQL system for Electronic Health Records that translates natural language questions into SQL queries and abstains from unanswerable questions.