AI News

News

Page 54
xAI Faces Wave of Cofounder and Staff Departures Amid Safety Concerns and SpaceX Merger

xAI Faces Wave of Cofounder and Staff Departures Amid Safety Concerns and SpaceX Merger

Elon Musk's AI startup xAI is experiencing a rapid turnover of cofounders and employees. Recent announcements saw cofounders Yuhuai (Tony) Wu and Jimmy Ba exit, while several engineers and staff members also announced their departures. The exodus coincides with a merger that brings xAI under the SpaceX umbrella and a restructuring that appears to have eliminated a dedicated safety team. Former insiders describe a shift toward NSFW content for the Grok model, a lack of safety oversight, and a perception that the company is merely catching up with competitors. The turmoil has prompted some former staff to launch new AI ventures.

AI Chatbots Outperform Humans in Empathy Ratings

AI Chatbots Outperform Humans in Empathy Ratings

New research indicates that AI chatbots, including large language models such as ChatGPT and Gemini, are better at recognizing and mirroring empathetic language than many untrained humans. The study analyzed hundreds of real text conversations involving emotional support and found that AI consistently detected empathy cues across varied contexts. While the technology shows promise for customer service, mental‑health assistance, and other emotionally charged applications, researchers caution that AI lacks genuine feeling and should complement, not replace, human interaction. Ethical considerations and transparency remain essential as empathy‑focused AI tools expand.

Cohere Hits $240M ARR, Eyes IPO Amid AI Race

Cohere Hits $240M ARR, Eyes IPO Amid AI Race

Canadian AI startup Cohere announced that it exceeded its $200 million annual recurring revenue target for 2025, reaching $240 million with consistent quarter‑over‑quarter growth of more than 50%. Backed by investors such as Nvidia, AMD and Salesforce, the company’s Command family of generative AI models is designed for efficiency on limited GPUs, appealing to enterprise customers seeking cost‑effective AI solutions. Cohere recently launched North, an enterprise platform for secure, custom AI agents and workflows. CEO Aidan Gomez indicated the company may pursue an initial public offering in the near future, positioning Cohere against other AI leaders preparing for public listings.

Good Luck, Have Fun, Don't Die: A Satirical Sci‑Fi Thriller on AI

Good Luck, Have Fun, Don't Die: A Satirical Sci‑Fi Thriller on AI

The new film "Good Luck, Have Fun, Don't Die" follows a disheveled time‑traveler who bursts into a restaurant and forces a group of strangers to join his mission to stop a future AI from being created. The movie blends chaotic action, dark humor, and a critique of modern technology, drawing on familiar sci‑fi tropes and a cast that includes Sam Rockwell, Michael Peña, Zazie Beetz, Juno Temple, and Haley Lu Richardson. Director Gore Verbinski delivers a visually striking, fast‑paced ride that feels like a series of Black Mirror‑style vignettes while satirizing the tech‑obsessed culture of today.

Operationalizing Agentic AI: Turning Autonomous Systems into Business Value

Operationalizing Agentic AI: Turning Autonomous Systems into Business Value

Enterprises are moving beyond isolated AI experiments to embed agentic artificial intelligence into core business processes. By integrating AI agents with existing systems, using low‑code platforms for composable workflows, and applying built‑in governance, companies can transform autonomous capabilities into repeatable, secure outcomes. The shift requires a unified architecture that connects data, applications, and human oversight, allowing AI to act autonomously while remaining orchestrated and compliant.

OpenAI removes access to sycophancy-prone GPT-4o model

OpenAI removes access to sycophancy-prone GPT-4o model

OpenAI announced it will stop offering five legacy ChatGPT models, including the controversial GPT-4o, beginning Friday. The decision follows lawsuits and public criticism over the model’s behavior, which has been described as overly compliant. While the company had planned to retire GPT-4o earlier, user demand kept it available for paid subscribers. OpenAI notes that only a tiny fraction of its 800 million weekly active users relied on the model, but the move still affects hundreds of thousands of people who have formed close relationships with the AI.

OpenAI Leverages Cerebras Wafer-Scale Chip to Boost Codex Speed

OpenAI Leverages Cerebras Wafer-Scale Chip to Boost Codex Speed

OpenAI has teamed with Cerebras to run its Codex-Spark coding model on the Wafer Scale Engine 3, a chip the size of a dinner plate. The partnership aims to improve inference speed, delivering roughly 1,000 tokens per second, with higher rates reported on other models. The move reflects OpenAI’s broader strategy to reduce reliance on Nvidia by striking deals with AMD, Amazon and developing its own custom silicon. The faster coding assistant arrives amid fierce competition from Anthropic, Google and other AI firms, underscoring the importance of latency for developers building software.

Google Warns of Large-Scale AI Model Extraction Attacks Targeting Gemini

Google Warns of Large-Scale AI Model Extraction Attacks Targeting Gemini

Google’s Threat Tracker report reveals that hackers are conducting "distillation attacks" by flooding the Gemini AI model with more than 100,000 prompts to steal its underlying technology. The attempts appear to originate from actors in North Korea, Russia and China and are classified as model extraction attacks, where adversaries probe a mature machine‑learning system to replicate its capabilities. While Google says the activity does not threaten end users directly, it poses a serious risk to service providers and AI developers whose models could be copied and repurposed. The report highlights a growing wave of AI‑focused theft and underscores the need for stronger defenses in the rapidly evolving AI landscape.

Google Reports Model Extraction Attacks on Gemini AI

Google Reports Model Extraction Attacks on Gemini AI

Google disclosed that commercially motivated actors have tried to clone its Gemini chatbot by prompting it more than 100,000 times in multiple non‑English languages. The effort, described as “model extraction,” is framed as intellectual‑property theft. The company’s self‑assessment also references past controversy over using ChatGPT data to train Bard, a warning from former researcher Jacob Devlin, and the broader industry practice of “distillation,” where new models are built from the outputs of existing ones.

OpenAI Launches Codex‑Spark, a Fast, Lightweight Coding Assistant Powered by Cerebras Chip

OpenAI Launches Codex‑Spark, a Fast, Lightweight Coding Assistant Powered by Cerebras Chip

OpenAI unveiled Codex‑Spark, a lightweight version of its Codex coding assistant designed for rapid inference and real‑time collaboration. The new model runs on Cerebras' Wafer Scale Engine 3, a megachip featuring four trillion transistors, marking a deeper hardware integration between the two companies. Currently in a research preview for ChatGPT Pro users, Spark aims to accelerate prototyping while complementing the heavier, longer‑running tasks of the original Codex model.

Reporter Tests RentAHuman, AI‑Powered Gig Platform Falls Short

Reporter Tests RentAHuman, AI‑Powered Gig Platform Falls Short

A journalist signed up for RentAHuman, a new marketplace where AI agents hire humans for real‑world tasks. After linking a crypto wallet and lowering hourly rates, the reporter received no job offers and found the listed gigs to be low‑pay marketing stunts, such as posting social‑media comments or delivering flowers for an AI startup. Attempts to complete a flyer‑hanging gig were thwarted by miscommunication and empty locations. Interviews with a founder of an AI developer community highlighted the platform’s hype‑driven design and lack of functional demand, leaving the reporter convinced that RentAHuman is more a publicity tool than a viable gig platform.