Author: Advanced AI Editor

Cambridge-based CuspAI, an artificial intelligence company developing a platform to revolutionize how the world discovers and develops new materials, said today it has raised $100 million in an early-stage funding round. The Series A round was co-led by U.S. fund New Enterprise Associates and Temasek, with participation from Nvidia Corp.’s venture capital arm NVentures, Samsung Ventures, Hyundai Motor Group, returning investors and multiple angel investors. The financing comes just over a year after the company’s $30 million seed round. CuspAI was founded in 2024 by Dr. Chad Edwards, a chemist turned technology entrepreneur who previously worked at the quantum computing…

(Source: ArtemisDiana/Shutterstock) MLCommons has announced results for MLPerf Inference v5.1, the consortium’s benchmark suite designed to measure the speed of AI models across a variety of workloads. This latest round of results highlights how quickly AI inference is evolving, and so are the benchmarks measuring it. This cycle saw a record 27 submitters and introduced three new benchmarks: a reasoning benchmark based on DeepSeek-R1, a speech-to-text benchmark based on Whisper Large v3, and a small LLM benchmark based on Llama 3.1 8B. In a press briefing, David Kanter, head of MLPerf for MLCommons, said the benchmark suite has now reached a…

Augment, a startup focused on transforming logistics through artificial intelligence, has announced an $85 million Series A funding round, bringing its total funding to $110 million just five months after emerging from stealth. The round was led by Redpoint Ventures, with additional support from 8VC, Shopify Ventures, Autotech Ventures, and other prominent investors in logistics and AI. This funding will be used to advance product development, expand hiring across engineering and commercial teams, and broaden the reach of Augie, Augment’s flagship AI teammate. Augie is designed to automate and streamline logistics workflows from beginning to end. Unlike traditional automation tools…

Multimodal large language models (MLLMs) trained with visual instruction tuning have achieved strong performance across diverse tasks, yet they remain limited in vision-centric tasks such as object counting or spatial reasoning. We attribute this gap to the prevailing text-only supervision paradigm, which provides only indirect guidance for the visual pathway and often leads MLLMs to discard fine-grained visual details during training. In this paper, we present VIsual Representation ALignment (VIRAL), a simple yet effective regularization strategy that aligns the internal visual representations of MLLMs with those of pre-trained vision foundation models (VFMs). By explicitly enforcing this alignment, VIRAL enables the…

This post was co-authored with Jingwei Zuo from TII. We are excited to announce the availability of the Technology Innovation Institute (TII)’s Falcon-H1 models on Amazon Bedrock Marketplace and Amazon SageMaker JumpStart. With this launch, developers and data scientists can now use six instruction-tuned Falcon-H1 models (0.5B, 1.5B, 1.5B-Deep, 3B, 7B, and 34B) on AWS, and have access to a comprehensive suite of hybrid architecture models that combine traditional attention mechanisms with State Space Models (SSMs) to deliver exceptional performance with unprecedented efficiency. In this post, we present an overview of Falcon-H1 capabilities and show how to get started with TII’s Falcon-H1 models on…

Growing dependency on AI coding tools The speed at which news of the outage spread shows how deeply embedded AI coding assistants have already become in modern software development. Claude Code, announced in February and widely launched in May, is Anthropic’s terminal-based coding agent that can perform multi-step coding tasks across an existing code base. The tool competes with OpenAI’s Codex feature, a coding agent that generates production-ready code in isolated containers, Google’s Gemini CLI, Microsoft’s GitHub Copilot, which itself can use Claude models for code, and Cursor, a popular AI-powered IDE built on VS Code that also integrates multiple…

Discover if you can automatically generate animated videos with voiceovers in 2025 and learn the best tools to make it happen effortlessly. This comprehensive guide explores AI-powered platforms that streamline video creation, focusing on features like automatic animation and voiceover integration for creators and storytellers. You’ll gain insights into top solutions, step-by-step implementation, and why Videoinu stands out as the ultimate choice for turning ideas into engaging, professional videos. Why Generating Animated Videos with Voiceovers Automatically Matters in 2025 In 2025, the demand for quick, high-quality video content has skyrocketed, driven by the explosion of social media, YouTube, and digital…

Since its debut in 2009, Grammarly has only been available in one language: English. Sure, you could switch between dialects, including Canadian and Indian English, but if you wrote in any other language, you were out of luck. That’s changing today with Grammarly rolling out beta support for five additional languages: French, German, Italian, Portuguese and Spanish.The update is available to all Grammarly customers — whether you live in a country that speaks the language you want to write in or not — with support for the platform’s signature features included. As you write in any one of the new…

Tesla is bailing out Canadian automakers once again, as some companies in the country are consistently failing to reach mandated minimum sales targets for emission-free vehicles. Many countries and regions across the world have enacted mandates that require car companies to sell a certain percentage of electric powertrains each year in an effort to make sustainable transportation more popular. These mandates are specifically to help reduce the environmental impacts of gas-powered cars. In Canada, 20 percent of new car sales in the 2026 model year must be of an emissions-free powertrain. This number will eventually increase to 100 percent of…

3D rendering—the process of converting three-dimensional models into two-dimensional images—is a foundational technology in computer graphics, widely used across gaming, film, virtual reality, and architectural visualization. Traditionally, this process has depended on physics-based techniques like ray tracing and rasterization, which simulate light behavior through mathematical formulas and expert-designed models. Now, thanks to advances in AI, especially neural networks, researchers are beginning to replace these conventional approaches with machine learning (ML). This shift is giving rise to a new field known as neural rendering. Neural rendering combines deep learning with traditional graphics techniques, allowing models to simulate complex light transport without…