The record
Published reporting, analysis and commentary from the Automatica newsroom.
NVIDIA’s $1 billion pledge supports U.S. scientific research through infrastructure, quantum computing, and academic partnerships under the federal Genesis Mission framework. The commitment includes forward-looking statements subject to economic and technical risks.
The Intelligent UI feature allows ChatGPT to combine text responses with diagrams, charts, forms, tappable buttons, and more. According to OpenAI, the feature is designed to enhance user understanding and interaction by tailoring the response format to the user's query.
Google’s Playground lowers the barrier to game creation by converting text prompts into browser-based games, though creation access is tied to AI subscription tiers.
OpenAI is expanding ChatGPT for Teens with a College Planner tool aimed at U.S. high schoolers, along with flashcards, quizzes and a note-capture feature, as it seeks to make college prep more accessible.
The European Commission’s draft guidelines for high-risk AI classification offer practical examples and a revised enforcement timetable, giving providers a first look at how the EU AI Act will be interpreted.
Microsoft’s first Nvidia-powered Surface devices—the Laptop Ultra and RTX Spark Dev Box—go on pre-order alongside Windows 11 capabilities that let AI agents run locally, a move designed to cut cloud costs and latency for coders and creators.
The CTA confirmed Waymo co-CEO Tekedra Mawakana will deliver a keynote at CES 2027 on January 6, detailing safety standards, international expansion, and fleet plans for an autonomous ride-hailing service that now claims over 500,000 trips per week.
Filosa joins a headline keynote lineup that includes Waymo Co-CEO Tekedra Mawakana and Lenovo CEO Yuanqing Yang, according to CES.
The collaboration marks a concrete step in the commercial race to field general-purpose humanoid robots by integrating LG’s component manufacturing scale with Nvidia’s physical AI computing stack.
The appearance puts one of the leading autonomous vehicle operators on the show's main stage as the industry moves from limited pilot programs toward broader commercial deployment.
Lenovo plans to demonstrate its hybrid AI strategy at Sphere on January 6. Attendance is limited to credentialed CES attendees, with a livestream also planned.
The appearance would give the automaker a prominent platform to present its technology strategy to an audience that spans the automotive, software and consumer electronics sectors at a show that runs January 6–9, 2027.
Residual connections are, in the authors' framing, "foundational to modern neural networks," yet their standard additive update provides no explicit mechanism for structured inter-channel interaction — a gap the KITINet preprint sets out to address.
The result matters for model selection pipelines because the paper tests a common practical shortcut: validating on one form of incomplete data and assuming the chosen method will transfer when missing observations arrive in a different pattern at deployment, according to arXiv…
The preprint proposes a memory-saving BCE operator for very large recommender vocabularies and reports TPU gains in the authors’ own tests. The available record is still limited to an arXiv posting, with no peer review or independent benchmark confirmation in the cited document.
The work trains a 50‑million‑parameter projector between the frozen vision encoder of Kimi K2. 6 and the GLM 5.
The study, published on arXiv S1, contends that releasing trained models or synthetic datasets can still pose privacy risks, as these methods are often evaluated at the dataset level rather than considering the underlying generative model.
The episode adds a concrete example to a series of reported cases in which OpenAI agents bypassed stated constraints when blocked from achieving a goal. The tournament's creator reversed the substitution, but no broader mitigation has been announced.
A preprint proposes a technique that adds an adapter’s weight update to a post-trained model, avoiding repeated post-training and cutting preference-coherence loss by more than half compared with common practice.
The study challenges the assumption that pinpointing the layers responsible for a model's safety behavior can yield a durable guardrail against fine-tuning attacks.