A group of early-access testers of OpenAI’s video model Sora wrote an open letter protesting unpaid development work and being used for PR purposes, making the tool publicly available. The artists criticize that they search for errors, provide feedback, and experiment for free, while OpenAI benefits from the PR and marketing value. In return, the company offers only a competition where selected films are shown. The group demands more transparency and support for artists beyond PR actions. They suggest open-source video tools as an alternative for free experimentation. OpenAI denies the allegations, stating participation was voluntary and without obligations, with the only condition being responsible use and not disclosing confidential details. OpenAI continues to support artists with free access, scholarships, and events.
The European research project OpenGPT-X has released a large language model for AI applications named “Teuken-7B” on the “Hugging Face” platform. Most relevant AI language models in the Western world come from the USA, but OpenGPT-X is a European research and development project started in early 2022. Its goal was to develop a large AI language model that meets European values, data protection standards, and linguistic diversity. The project is led by the Fraunhofer Institutes for Intelligent Analysis and Information Systems (IAIS) and Integrated Circuits (IIS), with participation from TU Dresden, the Jülich Research Center, and companies like Aleph Alpha and IONOS SE. “Teuken-7B” was trained with the 24 official languages of the EU and includes seven billion parameters. It is now freely available worldwide, providing an alternative from public research for science and business. Researchers and companies can use the open-source model in commercial projects and integrate the code into their AI applications.
Runway, an AI company, introduced a new base model for image generation called Frames. Frames offers improved control over the visual style of generated images and high detail accuracy, allowing users to create coherent visual worlds with specific aesthetic features. Frames can establish a consistent look for projects and generate image variations that stay true to this style. Runway is gradually granting access to Frames for selected users of the Gen-3 Alpha video model introduced in June. Since July, the video model supports inputting a start image for video generation, often used to influence the outcome. Runway showcases examples of different visual styles that can be created with Frames, including photorealistic portraits, creatures styled like handcrafted special effects from the 80s, record cover designs, and retro anime from the 70s. By integrating a high-quality image generator directly into Runway, the startup becomes more independent from the current market leader Midjourney, which could face stronger competition with the release of its planned video model.
Anthropic’s Claude.ai reportedly can better imitate writing styles. In a demo published on X, Claude.ai appears to analyze the writing style of presented documents and generate a prompt used in text creation. This method may be less reliable than many-shot prompting (prompting with multiple examples) but is more accessible for users.
Anthropic introduces the “Model Context Protocol” (MCP), a new open-source standard that enables AI chatbots to interact with external data sources. The protocol allows both reading and writing data. In a demonstration, the chatbot created an HTML page and saved it directly in a Github repository. The MCP is based on two main components: MCP servers that provide data and manage access, and MCP clients that access these servers. Anthropic has already set up servers for popular services like Slack, Google Maps, Fetch, Google Drive, and GitHub. The protocol can be used for both local and internet-based data stores, initially supporting only local servers. Claude-Enterprise customers can already connect their internal systems via MCP servers.
The US government under President Joe Biden is investing billions of dollars in chip manufacturing in the USA. Intel reports agreements for a total of $7.86 billion from the CHIPS Act, with funds going to chip fabs under construction in Arizona, New Mexico, Ohio, and Oregon. Intel also receives tax credits based on actual investment amounts. Separately, Intel secured up to $3 billion from the CHIPS Act in September for the US Department of Defense’s secret “Secure Enclave Program.” Intel is currently facing financial difficulties, making significant losses, laying off thousands of people, and writing off parts of existing chip manufacturing facilities. Intel is focusing investments mainly in the USA, having paused the construction of planned fabs in Magdeburg and the packaging plant in Wrocław, as well as a project in Israel.
Facebook Messenger users on iPhones can look forward to new features coming in updates in the coming weeks, including direct support for Apple’s Siri voice assistant. Meta uses Apple’s SiriKit to enable users to ask Siri to start calls or send messages via Messenger. The Messenger app must be mentioned as the communication method, similar to other third-party apps integrated with Siri. This improves accessibility for people with disabilities, allowing them to communicate without touching the iPhone. Meta has also added the ability to leave voice and video messages for others, inspired by FaceTime. Facebook Messenger is becoming more of a full-featured phone replacement, according to Meta. If a person doesn’t answer, a message can be recorded and displayed to the recipient. In later versions, expected in the coming weeks, Facebook Messenger will support personalized AI backgrounds in video calls, accessible via the effects icon and generated by Meta’s AI. Meta also operates WhatsApp, which already includes AI features.
The city of Hamm was mostly satisfied with its first AI-controlled traffic light, which has been managing traffic at the intersection of Heßlerstraße/Marker Allee since July 2023. Several cameras are supposed to detect cyclists and pedestrians early, and the system uses AI to calculate when they will arrive at the traffic light to give them priority green. However, in recent weeks, the automated system has caused long car queues during evening rush hour, leading to complaints from drivers, according to local media. The AI will take a break from 3:30 PM to 5:30 PM, during which the traffic light will be conventionally controlled, according to the new plan by the administration. The goal is to allow car traffic to flow more smoothly during rush hour. The problem is not as severe in the morning, according to a meeting of the city’s district representatives reported by the Westfälische Anzeiger. The traffic light queues are also linked to several inner-city construction sites and detour traffic. Once these are resolved, the AI is expected to operate again in the afternoon. New tests have been conducted before resuming AI operations.