Sarah Nakamura
AI News writer at AIxploria
Sarah Nakamura writes about AI research breakthroughs, benchmarks, and technical advances for AIxploria.
Topics covered
anthropic
meta
chatgpt
openai
muse spark
Articles by Sarah Nakamura (35)
- Three Days After Google Mapped a Fly Brain, It Plays Doom Google mapped all 166,700 neurons of a male fruit fly. Three days later, coders made the digital connectome play Doom, Mario 64 and Beat Saber. Yes, wild.
- From $0 to $100, Muse Plays AI Butler Inside WhatsApp Meta's Muse agent lives inside WhatsApp with its own virtual computer: it emails, books travel and shops for you. Free tier, then $20 or $100 a month.
- DLSS 5 Lands September 3, and Gamers Already Want It Gone DLSS 5 launches September 3 with NBA 2K27, exclusive to RTX 50 GPUs. Waxy faces, memes and backlash: Nvidia's AI graphics tech divides before day one.
- Beijing's Robots Outran Bolt, Then Face-Planted Into Foam A humanoid robot ran the 100m in 8.64s in Beijing, under Bolt's record. But precision events with tweezers and hotel beds tell the real story of robotics.
- In Ohio, marrying ChatGPT could be banned Ohio's HB 469 would void human-AI marriages and deny chatbots any legal personhood. Symbolic weddings with ChatGPT companions are already real in Japan.
- This Toyota Yaris Vanished From an AI Surveillance Camera An AI-generated pattern made a Toyota Yaris invisible to Flock surveillance cameras during a live Def Con test. Here's how it worked.
- MiniMax H3 Brings Open Weights to 2K AI Video, Audio Included MiniMax H3 goes open weight: 2K AI video, native stereo audio, 15-second clips at a fraction of rivals' prices. Veo and Sora just got a Shanghai problem.
- Apple Glasses With No Camera? The Anti-Meta Bet for 2027 Apple has reportedly pushed its smart glasses to 2027 and may drop the camera entirely. Privacy comes first, even if it kills the category's best trick.
- ChatGPT Will Pick Your Restaurant With Yelp's 330M Reviews Yelp is licensing 330 million reviews to OpenAI: ChatGPT will now recommend restaurants and request quotes for you. Google's local moat just got thinner.
- Meta's StoryKit Turns Your Kid Into a Storybook Hero Meta quietly launched StoryKit, an AI app that turns your child's photo into an illustrated storybook hero, complete with narration and a printable PDF.
- Kimi K3 vs. GPT and Claude: Can This Open-Weight Giant Really Win? Moonshot AI's Kimi K3 packs 2.8 trillion parameters, the biggest open-source AI model ever, undercutting Claude and GPT-5.6 on price and context.
- Emergent Hits $1.5B Unicorn Status in Just 13 Months Indian vibe-coding startup Emergent just became a $1.5B unicorn with a $130M Series C, only 13 months after launching its AI coding platform.
- 44% of New Songs Are AI. The Industry Just Labeled Them RIAA and IFPI unveil voluntary AI music labels as Deezer reports 44% of new uploads are already AI-generated tracks.
- FL Studio 2026: Gopher AI Can Now Touch Your Mixer FL Studio 2026 turns Gopher from a Q&A chatbot into an assistant that routes your mixer and sets levels for real.
- World Cup 2026: Ref Cams, 3D Avatars, and AI Analysts Are Rewriting How Football Works Ref cams, 3D avatars, Football AI Pro: the 2026 World Cup deploys the most ambitious AI toolkit football has ever seen on the pitch.
- Midjourney Medical: The AI Image Company Just Unveiled a Full-Body Scanner, and Yes, It's Serious Midjourney Medical unveils a full-body ultrasonic scanner: 60-second scans, no radiation, and a spa in San Francisco. Bold vision, but no FDA nod yet.
- Android 17 Hits Pixel Phones Today: Gemini Intelligence Is Here, but Your Phone Probably Can't Run It Android 17 launches on Pixel with Gemini Intelligence, but 12 GB RAM and Nano v3 requirements lock out most phones. Here's what you need to know.
- The US Government Just Pulled the Plug on Claude Fable 5 for Everyone The US government forced Anthropic to disable Claude Fable 5 and Mythos 5 via an export control directive. Here's what happened and why it matters.
- Claude Fable 5: Anthropic Releases Its Most Powerful Model to the Public, With a Safety Leash Anthropic launches Claude Fable 5, its Mythos-class model now available to the public with safety guardrails, doubled pricing, and 30-day data retention.
- Anthropic Wants a 'Brake Pedal' for AI Before It's Too Late: Claude Already Writes 80% of Its Own Code Anthropic warns AI is nearing recursive self-improvement, with Claude writing 80% of its own code. Brake pedal plea or pre-IPO move?
- AI Is Cracking Animal Language Across Six Species: The Dr. Dolittle Machine Is Real AI neural networks are decoding animal communication in mice, dolphins, whales and more. The Coller Dolittle Prize funds the race to talk to animals.
- ChatGPT Now Searches for Jobs and Formats Your Resume: LinkedIn, We Need to Talk ChatGPT adds live job search and resume formatting. OpenAI now pulls listings from Indeed and Upwork while letting you build polished CVs.
- DuckDuckGo Installs Spike 30% as Users Flee Google's AI-Heavy Search Overhaul DuckDuckGo installs spiked 30% in the U.S. after Google I/O's AI Search overhaul. Users want choice, not forced AI in every query.
- Spotify and Universal Just Made AI Remixes Legal: Fans Create, Artists Get Paid Spotify partners with Universal Music for AI-powered fan remixes and covers. A paid Premium add-on where artists earn royalties. Here's what it means.
- Google's UCP Buy Button Hits the Main SERP: Is Your Store Ready for Agentic Commerce? Google's UCP Buy button now appears in main search results. Learn how Semrush One's Site Audit and SEO tools prepare your structured data for agentic commerce.
- OpenAI's Daybreak Takes Aim at Anthropic's Glasswing in the AI Cybersecurity Arms Race OpenAI launches Daybreak, an AI cybersecurity platform powered by GPT-5.5, to rival Anthropic's Glasswing. Partners, tiered models, and what it means.
- China Just Made It Illegal to Fire Workers for AI Replacement, and No Western Country Has Anything Like It A Chinese court ruled firing workers to replace them with AI is illegal. No Western country has an equivalent legal protection yet.
- Meta Snaps Up Assured Robot Intelligence to Build the Android of Humanoid Robots Meta acquired Assured Robot Intelligence to build the Android of humanoid robots. Here's why this small deal signals a massive strategic shift for AI.
- Snapchat's AI Sponsored Snaps Put Brand Chatbots Right in Your DMs Snapchat rolls out AI Sponsored Snaps, placing brand-controlled AI chatbots in your Chat tab. Experian is the first partner in this bold ad play.
- Workspace Agents: OpenAI Just Turned ChatGPT Into Your Team's Always-On Digital Coworker OpenAI launches Workspace Agents in ChatGPT: Codex-powered autonomous agents that automate team workflows. Free until May 6, 2026, then credit-based.
- Mozilla's Thunderbolt: An Open-Source AI Client You Host Yourself, and It Actually Makes Sense Mozilla's MZLA subsidiary launches Thunderbolt, an open-source self-hosted AI client challenging Copilot and ChatGPT Enterprise on data sovereignty.
- Claude Design: Anthropic Wants You to Build Prototypes and Slides Without Opening Figma Anthropic launches Claude Design to generate prototypes, slides, and mockups from conversation. Powered by Opus 4.7, now in research preview.
- Canva AI 2.0: The Design Tool That Wants to Run Your Entire Creative Workflow Canva AI 2.0 launches in research preview with an agentic assistant, living memory, and deep integrations. Here's what it means for creators.
- Microsoft Bets Big on OpenClaw to Turn Copilot Into a Real Autonomous Agent Microsoft is integrating OpenClaw into Copilot 365 to build autonomous agents. Security risks, competition, and what it means for you.
- Muse Spark: Meta Spent $14 Billion to Catch Up in AI, and Ditched Open Source Along the Way Meta's Muse Spark AI model scores top 5 globally but ditches open source. Here's what it does, how it compares, and why privacy matters.
