# forck.live — 1–7 July 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-W27, 1–7 July 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-W27 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### Vercel — Understanding Vercel Functions - Company: Vercel (vercel.com) - Announced: 2024-07-05T13:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/understanding-vercel-functions - Record: https://forck.live/items/20969-understanding-vercel-functions - Subject: v0 / AI SDK / AI Gateway Vercel Functions run code in response to user traffic without the need to manage your own infrastructure, provision servers, or manage hardware. ### LG AI Research — [ICLR 2024]The Latest Research Trends of Foundation Model - Company: LG AI Research (lgresearch.ai) - Announced: 2024-07-05 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.lgresearch.ai/blog/view?seq=452 - Record: https://forck.live/items/4843-iclr-2024-the-latest-research-trends-of-foundation-model - Subject: EXAONE LG AI Research summarizes their contributions and the latest research trends on foundation models presented at ICLR 2024, focusing on understandability, sustainability, and safety of AI models. ### Vercel — Function streaming to be framework-agnostic on Vercel - Company: Vercel (vercel.com) - Announced: 2024-07-04T13:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://vercel.com/blog/vercel-functions-streaming-to-be-framework-agnostic - Record: https://forck.live/items/20970-function-streaming-to-be-framework-agnostic-on-vercel - Subject: v0 / AI SDK / AI Gateway Vercel announced that streaming HTTP responses will be enabled by default for all Node.js Vercel Functions. The rollout begins for Hobby accounts on July 8, 2024, and for Pro and Enterprise accounts on October 1, 2024. With streaming, responses and logs are sent in near real-time, and runtime errors after initial headers are sent result in a closed stream rather than a 500 error. ### Supabase — Postgres Realtime location sharing with MapLibre - Company: Supabase (supabase.com) - Announced: 2024-07-04 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://supabase.com/blog/postgres-realtime-location-sharing-with-maplibre - Record: https://forck.live/items/21421-postgres-realtime-location-sharing-with-maplibre - Subject: Supabase Supabase published a tutorial showing how to build a realtime location sharing application using PostGIS, Supabase Realtime, and MapLibre GL JS. The guide covers creating a Telegram Bot via a Supabase Edge Function to capture live location data, inserting it into Postgres with an RPC, and using Realtime to broadcast changes to React clients rendering markers on a map. ### Kyutai — Meet Moshi, the first real-time voice AI - Company: Kyutai (kyutai.org) - Announced: 2024-07-03 - Category: new-model - Coverage: not counted - Announcement: yes - Group: models - Source: https://kyutai.org/blog/2024-07-03-meet-moshi - Record: https://forck.live/items/18361-meet-moshi-the-first-real-time-voice-ai - Subject: Moshi / Pocket TTS Kyutai introduced Moshi, a real-time voice AI designed to enable expressive and spontaneous spoken interaction with machines. ### LG AI Research — [SNUxLG] Discovering the Infinite Potential of Future AI Leaders - Company: LG AI Research (lgresearch.ai) - Announced: 2024-07-03 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.lgresearch.ai/news/view?seq=450 - Record: https://forck.live/items/11024-snuxlg-discovering-the-infinite-potential-of-future-ai-leaders - Subject: EXAONE LG AI Research and the SNU-LG AI Research Center held an AI Tech & Talent Conference at Seoul National University, presenting joint research on 3D scene generation, conversation systems, and causal discovery, and offering recruitment consultations to over 200 AI students and researchers. ### LG AI Research — [SNUxLG] Discovering the Infinite Potential of Future AI Leaders - Company: LG AI Research (lgresearch.ai) - Announced: 2024-07-03 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.lgresearch.ai/blog/view?seq=450 - Record: https://forck.live/items/4844-snuxlg-discovering-the-infinite-potential-of-future-ai-leaders - Subject: EXAONE LG AI Research and Seoul National University held the AI Tech & Talent Conference, presenting joint research achievements, recruitment booths, and a poster session for AI talents. ### Hugging Face — Accelerating Protein Language Model ProtST on Intel Gaudi 2 - Company: Hugging Face (huggingface.co) - Announced: 2024-07-03 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/intel-protein-language-model-protst - Record: https://forck.live/items/1862-accelerating-protein-language-model-protst-on-intel-gaudi-2 - Subject: Platform To run inference and fine-tuning of the ProtST protein language model on Intel Gaudi 2 accelerators using the Optimum for Intel Gaudi library, achieving 1.76x faster inference and 2.92x faster fine-tuning compared to NVIDIA A100. ### Replit — Improved Dependency Management - Company: Replit (replit.com) - Announced: 2024-07-01T16:18:22+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://replit.com/blog/improved-dependency-management - Record: https://forck.live/items/17993-improved-dependency-management - Subject: Replit Agent Replit updated its workspace to consolidate language support, packages, and system dependencies into a single Dependencies pane. Users can now manage multiple languages in one Repl and install native programs or libraries via Nix packages under the System tab. The change aims to simplify configuration for both beginners and experienced developers. ### Upstage — Building end-to-end RAG system using Solar LLM and MongoDB Atlas - Company: Upstage (upstage.ai) - Announced: 2024-07-01 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.upstage.ai/blog/en/building-rag-system-using-solar-llm-and-mongodb-atlas - Record: https://forck.live/items/18480-building-end-to-end-rag-system-using-solar-llm-and-mongodb-atlas - Subject: Solar Pro / Solar Mini Upstage published a guide explaining how enterprises can build an internal RAG application using Solar LLM and MongoDB Atlas. The system retrieves information from sources like Google Workspace, Notion, GitHub, and Linear, and generates answers via a chatbot interface. The guide references Upstage's internal application 'Kim Solar' as an example. ### Suno — Introducing Suno for Mobile - Company: Suno (suno.com) - Announced: 2024-07-01 - Category: product-launch - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://suno.com/blog/suno-for-mobile - Record: https://forck.live/items/10348-introducing-suno-for-mobile - Subject: Music / Studio Suno launched a mobile app for iOS in the US, enabling users to make songs from text or recorded audio, listen to music from other creators, and curate playlists on their phones. ### LG AI Research — [NAACL 2024 Social Impact Award] Understanding the Capabilities and Limitations of Large Language Models for Cultural Common sense - Company: LG AI Research (lgresearch.ai) - Announced: 2024-07-01 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://www.lgresearch.ai/blog/view?seq=447 - Record: https://forck.live/items/4845-naacl-2024-social-impact-award-understanding-the-capabilities-and-limitations - Subject: EXAONE LG AI Research and the University of Michigan won the NAACL 2024 Social Impact Award for a paper that systematically analyzes the capabilities and limitations of large language models in understanding cultural common sense across five countries and languages. ### Hugging Face — Our Transformers Code Agent beats the GAIA benchmark 🏅 - Company: Hugging Face (huggingface.co) - Announced: 2024-07-01 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/beating-gaia - Record: https://forck.live/items/1863-our-transformers-code-agent-beats-the-gaia-benchmark - Subject: Platform Hugging Face's Transformers Code Agent achieved top performance on the GAIA benchmark, using a code agent approach with tools for web browsing and file inspection. The post also announces the upgrade of the Transformers Agents library to the standalone smolagents library.