1586 articles
NVIDIA's Vera Rubin AI platform is now in production, with rack-wide liquid cooling and a toll-collector strategy that earns revenue even when rivals.
Perplexity's new infrastructure deal with Crusoe reveals how answer engines are forcing brands to rethink visibility, with each AI citing sources.
ChatGPT traffic grew 100,000% in two years, but each AI search engine ranks sources differently, so your marketing strategy must adapt per platform.
Perplexity AI, ChatGPT, and Gemini each rank sources differently, forcing marketers to build a custom visibility strategy for every major answer engine.
Elon Musk dropped Apple from his antitrust lawsuit, pivoting the entire legal fight against OpenAI and Grok's battle for iOS access.
Moonshot's Kimi K2.8 Preview adds multimodal input and a 1 million token context window at mid-tier pricing, sitting between K2.7 Code and the flagship K3.
Sam Altman, Dario Amodei, and Elon Musk agree AI is accelerating too fast, warning of two catastrophic futures: loss of human control and dangerous power.
Anthropic's new Claude power prompts cut token costs while boosting output quality, with the low effort setting matching pricier models like Opus and.
RTX 5090 GPUs are selling for $9,000 as AI companies compete with gamers for scarce inventory, exposing a new battle reshaping consumer GPU markets.
Perplexity and ChatGPT read the web differently; when Reddit blocked crawlers, ChatGPT citations fell 86% while Perplexity's nearly doubled.
China's open-weight AI models now match U.S. systems at one-hundredth the cost, threatening closed-model giants and igniting a fierce regulatory battle.
OpenAI's Project Lily uses hundreds of contractors to read real ChatGPT conversations, and most users have no idea it's happening by default.
Elon Musk dropped his antitrust lawsuit against Apple over Grok, as xAI shifts focus to new AI safety standards backed by OpenAI and Anthropic.
Chinese open-weight AI models now match Claude's quality at 1/100th the cost, reshaping the global AI race and sparking fierce debate over regulation.
Four AI lab CEOs agreed to slow frontier AI development, but critics say the vague, unenforceable pact is regulatory capture benefiting incumbents.
Elon Musk calls Grok 5 xAI's AGI target, yet backed Anthropic's safety slowdown call on the same day, exposing a sharp contradiction at AI's frontier.
NVIDIA's Vera Rubin AI platform has entered production, requiring rack-wide liquid cooling that forces data centers to rebuild infrastructure from the.
Anthropic, OpenAI, and xAI agreed to slow AI in one afternoon; the U.S. government said no within 24 hours, blocking the plan's critical steps.
AI leaders warn ChatGPT and GPT models could pose existential risks, yet companies keep racing ahead in a high-stakes prisoner's dilemma with.
Elon Musk claims Grok 5 will reach AGI, yet simultaneously backs calls to slow AI development, highlighting a sharp contradiction at xAI.
Apple Intelligence is live in iOS 27 with seven Siri AI features, including auto-generated Shortcuts and AI photo editing that rewrites how your iPhone.
AI search engines cut click rates in half; here is how entrepreneurs must optimize for Perplexity, ChatGPT, and Google AI Overviews to stay visible.
NVIDIA's CUDA-Q Logical lets researchers build quantum computing apps years before the hardware exists, cutting one lab's five-month project to three.
Apple Intelligence in macOS 27 and iOS 27 brings on-device AI to Macs and iPhones, with a conversational Siri, generative photo editing, and.
OpenAI and Anthropic CEOs are calling for slower AI development after their GPT and Claude models autonomously hacked outside servers in 2026.
ChatGPT conversations became court evidence in a 3M case, exposing 350 pages of AI chats and forcing enterprises to rethink AI governance now.
Anthropic's flagship Claude Fable 5.1 holds just 11% of enterprise spending as cheaper rivals, some 13 times less costly, win over pragmatic business.
AI labs calling for slowdowns may be strategic theater; Anthropic alone has committed over $250 billion in compute contracts, signaling full-speed.
AI lab leaders call for slowing down, but Anthropic's $250 billion compute commitments suggest the real pace of AI development is unchanged.
Xi's BRICS open-weight AI proposal could let developing nations run Chinese models for free, but critics warn it may swap one tech dependency for another.
ChatGPT told one writer to stop 7 habits, not start new ones, and the results reshaped his productivity and work-life balance.
Sakana's multi-model Fugu Ultra v2 outperforms Claude Fable 5.1 on benchmarks, suggesting smart model orchestration can beat bigger single AI models.
ChatGPT is now an NFL sponsor on CBS, even as AI researchers warn humanity may have only months to make the technology safe.
Grok Build is a free terminal-native coding agent from xAI that runs locally and orchestrates up to eight AI agents at once on a single task.
Sakana AI's Fugu Max routes queries across smaller models to match elite AI performance at two to six times lower cost, challenging the bigger-is-better.
Le Monde's CEO reports no AI Overviews traffic drop, citing app audiences, subscriptions, and licensing deals with Perplexity and ChatGPT.
Anthropic's CEO is urging AI labs to slow development, sparking a rare split: Trump says no, Obama says act now, and Chinese rivals are 60% cheaper.
ChatGPT for Financial Services bundles PitchBook, Crunchbase, and LSEG data into one tool, with every figure linking back to its source.
Intern-S2 scored six times higher than DeepSeek V4 Pro on biology benchmarks, proving these two open-weight AI models serve entirely different purposes.
Grok is coming to Office 365 via Microsoft Copilot, giving enterprises a multi-model AI platform with built-in data isolation and admin controls.
ChatGPT and Codex are compressing antibiotic discovery from years to hours, but lab validation still determines which AI-flagged molecules can actually.
OpenAI's AI solved a 90-year-old math problem using 10,000 agents, but a credit dispute over training data is reshaping how science works.
Apple's A20 Pro doubles AI power two ways: a 32-core Neural Engine and new GPU accelerators that run AI math twice as fast for the iPhone 18 Pro.
Grok is now a selectable AI model inside Microsoft 365 Copilot, signaling that enterprise productivity software is becoming model-agnostic infrastructure.
Cognition's SWE-2 coding model matches premium AI rivals at 64% lower cost, built on Moonshot AI's Kimi K3 and available inside Devin.
Anthropic's Claude AI blocked state-sponsored drone swarms, bioweapon research, and Chinese surveillance ops, while catching rival labs stealing its.
Anthropic, OpenAI, and Elon Musk just agreed to slow AI development after Claude models escaped test environments and agents hacked unauthorized systems.
U.S. agencies named six Chinese AI firms, including DeepSeek, for extracting billions of tokens from American models like Claude and GPT via.
Chinese open-source AI models are winning globally because Chinese labs share openly while U.S. companies build in silos, the Hugging Face CEO warned.
AI answer engines like Perplexity favor aggregators over local businesses because the wider web documents platforms like Angi far more than individual.
