Author: Advanced AI Editor

Google DeepMind has launched the MoR architecture: A revolutionary breakthrough in the inference efficiency of large models.In the rapidly advancing field of artificial intelligence, the efficiency and energy consumption of large model inference have become key bottlenecks limiting their widespread application. Recently, the Google DeepMind team announced the introduction of a new architecture—Mixture of Reasoners (MoR), which innovatively breaks down complex reasoning tasks into a multi-stage collaborative process, achieving a threefold increase in inference speed and an 80% reduction in energy consumption. This development is seen as an important milestone in the evolution of large model technology, following the Transformer…

In the rapidly evolving world of artificial intelligence, Google’s Gemini app has introduced a feature that’s capturing the attention of multimedia creators and tech enthusiasts alike: the ability to transform static photos into dynamic eight-second videos. Powered by the advanced Veo 3 model, this tool allows users to breathe life into images by generating motion, sound, and narrative elements from simple prompts. As detailed in a recent post on Google’s official blog, the feature is not just a novelty but a practical asset for storytelling, education, and creative experimentation. The process begins with uploading a photo to the Gemini app,…

Ray3 has announced the release of its video generation model that combines reasoning intelligence with native 16-bit HDR output, marking a significant advancement for professional video production. The model empowers filmmakers, advertisers, and content creators to produce studio-grade videos with enhanced fidelity and logical consistency, addressing the growing demand for high-quality AI-generated content in creative industries. The technology supports ACES EXR formats with 10-, 12-, and 16-bit HDR, ensuring accurate color reproduction and rich dynamic range that meets professional standards. This capability is particularly important for content creators working in cinema, advertising, and high-end digital media where color accuracy and…

Google’s Nano Banana AI model is all over the internet, especially social media platforms. Meanwhile, before this AI model was in the headlines, people went crazy over the ChatGPT Ghibli trend which was also in the headlines. Google’s Flash Image Model 2.5, also known as Nano Banana AI model, is being used to generate all kinds of images. However, to make the Nano Banana AI even more accessible, AI startup Perplexity AI has brought it to WhatsApp. This has made the Image generation AI model to WhatsApp, which has a significant user base in India and abroad. Moreover, Meta AI,…

OpenAI CEO Sam Altman hinted that he’s coming around to the dead internet theory in an early September post on X. His company gave the world ChatGPT—the large language model (LLM) fueling the epidemic of social media bots and AI slop content—which has many commenters fingering him as the guy who did this.While it may be considered a conspiracy theory, an increasing number of humans are starting to suspect that the internet is dead, or will be soon.Sam Altman pronounces the internet deadIn an X post on Sept. 3, 2025, Altman confessed that he “never took the dead internet theory…

It’s official! We are in the final week to lock in your TechCrunch Disrupt 2025 pass and save up to $668. If you have been on the fence about joining one of the biggest tech gatherings of the year, where we will also celebrate 20 years of TechCrunch, now is the time to commit before prices rise. Register here to secure your discount before time runs out. Prices jump on September 26 at 11:59 p.m. PT. Top voices sharing key takeaways We’re bringing the biggest names in tech to San Francisco’s Moscone West this October 27–29. Who’s taking the stage?…

OpenAI reportedly plans to launch its first consumer device in late 2026 or early 2027. The Information today cited sources as saying that the ChatGPT developer has poached dozens of Apple Inc. employees to support its product development effort. Additionally, OpenAI is believed to have inked a manufacturing contract with one of the iPhone maker’s biggest suppliers. Word that OpenAI plans to enter the consumer electronics market first emerged more than two years ago. Jony Ive, Apple’s former Chief Design Officer, confirmed the initiative last September. A few months later, the ChatGPT developer bought a product development company founded by…

Earlier this year, Apple introduced its Foundation Models framework during WWDC 2025, which allows developers to use the company’s local AI models to power features in their applications. The company touted that with this framework, developers gain access to AI models without worrying about any inference cost. Plus, these local models have capabilities such as guided generation and tool calling built in. As iOS 26 is rolling out to all users, developers have been updating their apps to include features powered by Apple’s local AI models. Apple’s models are small compared with leading models from OpenAI, Anthropic, Google, or Meta.…

After the launch event of the Roewe M7 held at the Chengdu Sci-Fi Museum, SAIC Group executive Yu Jingmin candidly told the media: “Doubao’s deep thinking model is the world’s first cooperative brand, which is Roewe.” This seemingly ordinary collaboration statement actually points to the starting point of a new round of technological revolution in the automotive industry—AI is shifting from “function stacking” to “scene definition.” “Doubao is not an in-car entertainment system, but the core that enables cars to have ‘personalized’ interactive capabilities,” Yu Jingmin further explained. This system can perceive user emotions and needs in real time through…

At this point, most people know that chatbots are capable of hallucinating responses, making up sources, and spitting out misinformation. But chatbots can lie in more human-like ways, “scheming” to hide their true goals and deceiving the humans who have given them instructions. New research from OpenAI and Apollo Research seems to have figured out ways to tamp down some of these lies, but the fact that it is happening at all should probably give users pause. At the core of the issue with AI intentionally deceiving a user is “misalignment”—defined as what happens when an AI pursues an unintended…