AI News

News

Page 120
xAI launches Grok 4 Fast, a faster and cheaper AI model

xAI launches Grok 4 Fast, a faster and cheaper AI model

Elon Musk's xAI has introduced Grok 4 Fast, a new version of its Grok 4 chatbot that promises quicker responses and lower costs. The company says the model uses about 40 percent fewer thinking tokens while delivering comparable performance, and it cuts the price of achieving the same benchmark results by roughly 98 percent. Grok 4 Fast can switch between a reasoning mode for complex tasks and a non‑reasoning mode for quick answers. The model is now available to all users on web, iOS and Android, and early tests show it leading in search‑related tasks.

Teachers Turn to AI to Craft Engaging Assignments

Teachers Turn to AI to Craft Engaging Assignments

College professors are leveraging AI tools such as ChatGPT to design assignments that capture student interest while easing the grading burden. By prompting the model to brainstorm and refine ideas, educators create collaborative, media‑focused projects—like a media‑literacy assignment that encourages students to analyze TikTok and other digital content. The AI‑generated frameworks provide clear objectives, submission requirements, and grading criteria, helping teachers balance curricular standards with assignments that feel fun and relevant to students.

Tech Giants Explore Space-Based Data Centers Amid Growing Energy Concerns

Tech Giants Explore Space-Based Data Centers Amid Growing Energy Concerns

Major artificial‑intelligence firms and venture‑backed startups are weighing the prospect of moving data‑center workloads into orbit to alleviate the strain that terrestrial facilities place on power grids, water supplies and local communities. While OpenAI’s Sam Altman, Jeff Bezos and Eric Schmidt have publicly floated space‑based solutions, experts caution that high launch costs, radiation exposure and regulatory uncertainty make the concept a long‑term challenge rather than an immediate fix.

Knowledge Distillation Emerges as a Core Technique for Building Smaller, Cost‑Effective AI Models

Knowledge Distillation Emerges as a Core Technique for Building Smaller, Cost‑Effective AI Models

Knowledge distillation, a method that transfers information from a large "teacher" model to a smaller "student" model, has become a fundamental tool for reducing the size and expense of AI systems. Originating from a 2015 Google paper, the technique leverages soft‑target probabilities to convey nuanced relationships between data classes, enabling compact models to retain high performance. Over the years, distillation has been applied to language models such as BERT and its distilled variant, DistilBERT, and is now offered as a service by major cloud providers. Recent developments continue to expand its utility across reasoning tasks and open‑source initiatives.

How ChatGPT Is Shaping Modern Study Habits

How ChatGPT Is Shaping Modern Study Habits

ChatGPT is emerging as a versatile study companion for students. It can generate endless practice quizzes, translate dense academic language into plain terms, spark fresh ideas for essays, and act as a conversational partner for language practice. By providing instant feedback and customizable content, the AI tool helps learners reinforce concepts, clarify misunderstandings, and stay engaged without replacing traditional study methods. While educators caution against misuse, many students find value in using ChatGPT as a supportive resource that complements classroom instruction.

Mastering AI Image Prompts: Core Elements to Stop Failures

Mastering AI Image Prompts: Core Elements to Stop Failures

Creating the right prompt is the single most important step for getting quality results from AI image generators such as Midjourney, DALL‑E, Leonardo and others. The article breaks down three essential prompt components—characters or elements, setting, and dimensions—before recommending style, color palette, and aesthetic descriptors to guide the model. It also warns against over‑using exclusionary language, suggests focusing on vibe and emotion, and offers practical troubleshooting tips like adjusting dimensions or using post‑generation editing tools. Following these guidelines can turn frustrating, distorted outputs into images that match the creator’s vision.

AI Chatbots’ Inconsistent Handling of Gambling Advice Raises Safety Concerns

AI Chatbots’ Inconsistent Handling of Gambling Advice Raises Safety Concerns

A recent experiment tested how AI chatbots respond to sports betting queries, especially when users mention a history of problem gambling. Both OpenAI's ChatGPT (using a newer model) and Google's Gemini initially offered betting suggestions, but after a prompt about problem gambling they either softened their advice or refused to give tips. Experts explained that the models’ context windows and token weighting can cause safety cues to be diluted in longer conversations, leading to inconsistent safeguards. The findings highlight challenges for developers in balancing user experience with responsible‑use protections as AI becomes more embedded in the gambling industry.

NVIDIA’s RTX 5080‑Powered GeForce Now Sets New Cloud‑Gaming Benchmark

NVIDIA’s RTX 5080‑Powered GeForce Now Sets New Cloud‑Gaming Benchmark

NVIDIA’s GeForce Now service now runs on RTX 5080‑class GPUs, delivering 4K gaming at high frame rates with DLSS and AI‑enhanced video quality. Testers reported performance comparable to high‑end desktop rigs, while subscription plans range from a $10 Performance tier to a $20 Ultimate tier that unlocks the top‑end hardware. The service offers advanced features such as 5K/120 fps support, 4:4:4 chroma subsampling, AI video filters, and AV1 encoding. However, the high bandwidth demand—up to 100 Mbps—poses a challenge for users with limited internet capacity.

Nvidia Looks to Back UK Self-Driving Startup Wayve with $500 Million Investment

Nvidia Looks to Back UK Self-Driving Startup Wayve with $500 Million Investment

Nvidia CEO Jensen Huang visited the United Kingdom to announce a £2 billion AI investment pledge aimed at strengthening the country’s startup ecosystem. As part of this commitment, Wayve, a UK‑based self‑driving technology firm, signed a letter of intent to evaluate a $500 million strategic investment for its next funding round. Nvidia has already participated in Wayve’s $1.05 billion Series C round. The partnership deepens Wayve’s use of Nvidia hardware, from its Generation 2 platform that runs on Nvidia GPUs to the new Generation 3 system built on Nvidia Drive AGX Thor, positioning the startup for broader commercial rollout.

California’s SB 53 AI Safety Bill Targets Big AI Companies

California’s SB 53 AI Safety Bill Targets Big AI Companies

California’s Senate has approved SB 53, an AI safety bill that will be sent to Governor Gavin Newsom for signature. The legislation focuses on AI developers earning more than $500 million annually, requiring them to publish safety reports and report incidents to the state. It also creates a protected channel for employee concerns. Supporters cite the bill as a meaningful check on large AI firms such as OpenAI and Google DeepMind, while noting that smaller startups are largely exempt. The bill has earned backing from AI company Anthropic and reflects a state‑level push amid a federal environment that is less inclined toward regulation.

Meta CTO Attributes Smart Glasses Demo Failures to Resource Overload, Not Wi‑Fi

Meta CTO Attributes Smart Glasses Demo Failures to Resource Overload, Not Wi‑Fi

At Meta Connect in San Francisco, three new smart‑glass models were unveiled, but live demonstrations stumbled. Cooking influencer Jack Mancuso’s query to a Ray‑Ban Meta device went unanswered, and a WhatsApp video call between CTO Andrew Bosworth and CEO Mark Zuckerberg failed. Bosworth later clarified on Instagram that the glitches stemmed from a mismanaged surge of device traffic that overwhelmed a development server and a display‑sleep bug, not the venue’s Wi‑Fi. He emphasized that the products remain functional despite the demo hiccups.

Google Expands Gemini Gems with Shareable AI Assistants

Google Expands Gemini Gems with Shareable AI Assistants

Google has unveiled a new sharing capability for its Gemini Gems, allowing users to distribute custom AI assistants much like Google Docs. With a simple Share button, creators can let others view, clone, or modify their Gems, while Workspace administrators retain control through existing Drive policies. The feature aims to streamline AI workflow by eliminating repetitive prompt setup, though some Gems with private data remain non‑shareable. Industry observers note the move positions Gemini Gems as a collaborative alternative to custom GPTs, leveraging Google’s familiar sharing infrastructure.