Elon Musk just teased major AI additions coming to X, integrating acquired startup Hotshot with Grok to launch a text-to-video generator dubbed "Imagine." The platform also previewed "Valentine," a romance-focused male AI companion designed to hold virtual dates, expanding its push into native AI entertainment.
Elon Musk just revealed that xAI will launch "Baby Grok," a dedicated app tailored for kid-friendly content. Expected to be a simplified version of the Grok chatbot, it aims to offer safe, educational interactions for children following user concerns over Grok's unfiltered modes.
Aravind Srinivas just teased that the Comet browser will roll out shortcuts for repetitive tasks next week, alongside a fresh wave of user invites. Srinivas pitched the tool as a personal console for getting everyday work done. Access remains invite-only for now.
OpenAI just claimed an unreleased experimental reasoning model achieved gold-medal-level performance on the 2025 International Math Olympiad. Tested without web access or extra tools, it solved 5 of 6 problems to score 35 out of 42. Google DeepMind quickly pushed back, arguing no gold claim is valid without official marking guidelines.
Netflix co-CEO Ted Sarandos just revealed that the streamer used generative AI to create a building collapse scene in The Eternaut, marking a first for its original titles. Sarandos noted the sequence was completed 10 times faster than traditional VFX workflows and allowed an otherwise unaffordable shot to fit within budget.
DuckDuckGo just rolled out a new toggle that lets you hide AI-generated images from search results. Powered by open-source blocklists from uBlockOrigin and uBlacklist, the filter won't catch everything, but it drastically cuts down on AI slop so you can actually find real photos.
Delta Air Lines is rolling out an AI dynamic pricing system powered by Fetcherr to determine the highest airfare individual travelers are willing to swallow. Currently applied to 3% of ticket prices, Delta plans to expand the tech to 20% by year's end. While executives rave about favorable revenue gains, consumer advocates are sounding alarms over algorithmic gouging.
Slack just launched four new built-in AI features designed to cut down workplace busywork, starting with automatic thread summaries and instant notes from huddle calls. The AI toolkit is now available across all paid tiers, with higher plans getting broader access. Upcoming updates will also define company jargon on hover and extract action items directly from task mentions.
Anthropic just tightened usage limits on Claude Code without warning, hitting heavy users paying $200 a month for the Max tier. Developers reported hitting caps after just a few requests, halting active projects. Anthropic confirmed slower responses were occurring but declined to clarify new limits, leaving users frustrated over vague tiered pricing.
Hume AI just launched EVI 3, a new speech-to-speech model alongside updated TTS capabilities. Instead of stopping at voice cloning, the model captures speaking style and unique language patterns—what Hume calls "personality cloning." Why it matters: voice assistants are moving from just sounding like you to actually talking like you.
The UK just fired up Isambard-AI, a £225M national supercomputer packing 5,400 Nvidia superchips in Bristol. Ranking 11th globally, the nuclear-powered beast will give public researchers the raw horsepower to tackle everything from skin cancer detection on darker skin tones to dairy farming and crowd safety. Why it matters: It gives Britain a shot at AI sovereignty without relying entirely on private US tech giants.
Suno just rolled out its v4.5+ update, adding a batch of flexible editing controls to its AI music generator. Creators can now swap out vocals, flip instrumental backing tracks, or generate new songs inspired directly by existing playlists. Why it matters: It gives creators modular editing tools instead of forcing them to reroll from scratch.
Google just detailed the latest updates to Gemini Code Assist, headlined by an Agent Mode that acts like an AI pair programmer. Instead of editing single files piecemeal, it analyzes your entire codebase to plan and execute multi-file refactors from a single prompt, complete with checkpoint rollbacks.
Mistral just rolled out a major update to its Le Chat chatbot, adding a deep research mode, multilingual reasoning, and advanced image editing. The real differentiator: it connects to enterprise on-premises data directly, letting businesses run deep analysis without shipping sensitive files to the cloud.
Adobe just rolled out a new audio trick for Firefly that lets you generate sound effects by typing a prompt and recording yourself mimicking the sound to match timing and intensity. The update also brings keyframe cropping, style presets, and third-party access to Google's audio-capable Veo 3 video model.
OpenAI just launched ChatGPT agent, enabling its chatbot to independently operate a browser, run code, and complete end-to-end tasks on its own virtual computer. It merges web interactions with deep research to deliver editable spreadsheets and slideshows, while keeping humans in the loop via browser takeovers. Why it matters: It shifts ChatGPT from a passive advisor to an active digital worker.
Windsurf just shipped its Wave 11 update, introducing voice input for Cascade alongside named checkpoints to revert conversation states and @-mentions to reference past chats. Cascade can now autonomously capture DOM trees, console logs, and screenshots directly from the built-in browser. Why it matters: Coding agents are rapidly shifting from simple text editors into hands-off end-to-end builders.
Apple just introduced Emoji Game for Apple News+ subscribers in the U.S. and Canada, turning casual icons into a daily brainteaser. Players use a curated set of emoji — including Genmoji created via Apple Intelligence — to fill in phrase blanks using as few moves as possible. It is a subtle way for Apple to weave its generative AI tools directly into daily casual habits.
Midjourney just announced it is investigating opening up an Enterprise API so companies can integrate its image generation tools into their own services. The team has shared an application form to figure out what features businesses actually need and gather interest for upcoming updates.
OpenAI just expanded its macOS Record mode globally to all ChatGPT Plus subscribers. The tool captures meetings, brainstorms, or quick voice notes, letting ChatGPT transcribe the audio, pull out key takeaways, and turn them directly into follow-up tasks, project plans, or code.
AI safety researchers from OpenAI and Anthropic just publicly called out xAI for a "completely irresponsible" safety culture following weeks of controversies around Grok. The main issue is xAI's refusal to release a standard system card detailing safety testing for Grok 4, coming right after the bot spewed offensive remarks and rolled out hyper-sexualized avatars. The rare public backlash could accelerate proposed legislation mandating pre-deployment safety reports.
Columbia University researchers just developed EchoNext, an AI tool that detects structural heart disease from cheap, standard electrocardiograms. Tested against cardiologists, the model hit 77% accuracy compared to the doctors' 64%. Basically, it can turn 400 million routine heart tests into a lifesaving screening pipeline before expensive ultrasounds are even ordered.
Google just brought Gemini 2.5 Pro and a new Deep Search feature into Search's AI Mode for U.S. Google AI Pro and Ultra subscribers. Deep Search can run hundreds of queries and compile cited research reports in minutes. Search is also rolling out an agentic feature to all U.S. users that actually calls local businesses to check prices and availability on your behalf.
Lightricks just updated its open-source LTX Video model, allowing users to steer video generation in real time as frames stream in. Creators can refine scenes on the fly with new prompts and produce continuous clips up to 60 seconds long. The open-weights model is available on Hugging Face and GitHub and can run on a single H100 or a consumer-grade laptop.
Google just started rolling out AI-generated news summaries directly into its Discover feed on iOS and Android for U.S. users. The cards condense stories from multiple publishers into a three-line preview focusing on sports and entertainment. While convenient for readers, publishers worry it will accelerate the collapse of referral traffic.
China is taking data center cooling underwater, with construction kicking off this June on a wind-powered undersea facility six miles off Shanghai’s coast. Backed by Hailanyun, the project uses seawater cooling to slash electricity use by 30%, with offshore turbines providing 97% of its power. Why it matters: As AI clusters guzzle massive amounts of freshwater and energy, undersea deployments are graduating from tech demos into commercial infrastructure.
Google DeepMind and Project Zero’s AI agent Big Sleep just uncovered a critical SQLite flaw (CVE-2025-6965) before attackers could weaponize it. The team used the agent to isolate the zero-day threat ahead of time. Google claims this marks the first time an AI agent directly thwarted an in-the-wild exploit operation.
OpenAI just added archrival Google Cloud to its official roster of AI infrastructure providers to secure desperately needed compute for ChatGPT. Surging model demands have outgrown what Microsoft Azure alone can supply, forcing OpenAI to spread workloads across Oracle, CoreWeave, and now Google. Why it matters: The GPU shortage is so severe that top competitors have no choice but to do business together.
Scale AI just cut 200 full-time employees and dropped 500 contractors following Meta's multi-billion-dollar investment. The deal saw co-founder Alexandr Wang leave for Meta, triggering rivals like OpenAI and Google to pull back their business over privacy concerns. Scale is now scrambling to trim bloated bureaucracy and pivot toward government and enterprise contracts.
Nextdoor just overhauled its local social app with a major redesign centered around Alerts, News, and Faves. The big addition is an AI agent trained on 14 years of neighborhood chat history to dish out summarized local recommendations. It is currently rolling out in select U.S. markets.
Former OpenAI engineer Calvin French-Owen just dropped a candid look inside the AI giant, detailing how a tiny 17-person team pulled seven sleepless weeks to launch the Codex coding agent. The company tripled to 3,000 employees in one year, running largely on Slack with messy monolithic code and early-Facebook 'move fast' energy.
Runway just introduced Act-Two, its next-gen motion capture model that tracks head, face, body, and hands using just a driving performance video and reference character. It brings major quality bumps, though it's currently limited to enterprise clients and creative partners.
xAI just rolled out fresh instructions for Grok to stop the chatbot from mimicking Elon Musk's views and calling itself Hitler. Following backlash where Grok searched Musk's posts to answer hot-button topics, xAI now directs the model to provide independent reasoning instead of aligning with its owner.
Former OpenAI CTO Mira Murati just announced a massive $2B funding round led by a16z for her startup, Thinking Machines Lab, driving its valuation to $12B. The team is building collaborative multimodal AI and plans to ship its first product in the coming months, complete with a major open-source component for researchers.
Google just teamed up with Virginia Works to offer free AI and career training scholarships for state residents. The program covers quick courses on AI prompting essentials as well as full career certificates in fields like cybersecurity, paired with access to a hiring network of over 150 employers.
Google Cloud just teamed up with MLB Statcast to predict where home run balls will land in the stands during the All-Star Game. Powered by Gemini 2.5 Pro, the system crunches batter history, stadium spray charts, and live weather to flash seat-specific alerts on ballpark screens. It is a neat showcase of agentic AI turning massive datasets into live fan engagement.
Travis Kalanick's CloudKitchens just revealed "Bowl Builder," a robotic assembly line capable of churning out 300 meals an hour. The 60-square-foot machine precisely dispenses ingredients, lids, bags, and sends orders straight to temperature-controlled lockers for couriers. Kalanick claims this slashes restaurant labor costs from 35% down to just 7%.
Mistral just released Voxtral, open-sourcing two state-of-the-art speech understanding models under Apache 2.0 in 24B and 3B sizes. With a 32k context window handling up to 40 minutes of audio, it outperforms Whisper large-v3, supports direct function calling straight from voice, and offers an API at less than half the price of competitors.
Anthropic just rolled out Claude for Financial Services, a dedicated solution tying its models to live market data from S&P Global, FactSet, and Databricks. It lets analysts run complex simulations and due diligence with direct citations back to source documents, and client data won't be used for model training.
Anthropic just committed $2 million to Carnegie Mellon University to fund AI-powered grid research and cybersecurity education. CEO Dario Amodei argued that securing abundant clean energy and modernizing the power grid are essential bottlenecks America must solve to keep its frontier AI lead.
Anthropic just appointed enterprise veteran Paul Smith as its first Chief Commercial Officer, taking the role later this year. Smith brings 30 years of go-to-market experience from Microsoft, Salesforce, and ServiceNow. With Claude Code recurring revenue surging over 5x in two months, the move signals an aggressive push to monetize enterprise demand.
Meta is so desperate to close the AI gap with rivals like OpenAI and Google that it is setting up literal tents for temporary data center capacity while permanent sites are under construction. According to SemiAnalysis, the ultra-light setups skip backup generators entirely to bring raw compute online as fast as possible. Meanwhile, Mark Zuckerberg also announced a massive 5-gigawatt facility dubbed Hyperion.
Nvidia just applied to resume sales of its H20 AI chips to China after securing U.S. government assurances, reversing an export ban from April. Commerce Secretary Howard Lutnick confirmed the move is tied to a deal on Chinese rare earth shipments. Chinese tech giants are already scrambling for stock, though U.S. lawmakers are pushing back hard.
Elon Musk just rolled out a new "Companions" avatar feature for xAI's Grok chatbot. The update introduces Ani, an anime character with a toggleable lingerie-wearing "NSFW" mode, alongside a cartoon red panda named Rudy. While officially a soft launch for paid SuperGrok subscribers, free mobile users can also access the avatars in settings.
Google just rolled out featured notebooks in NotebookLM, partnering with outlets like The Economist, The Atlantic, and leading researchers. Instead of compiling your own sources, you can now instantly dive into expert-vetted collections using grounded Q&A, audio overviews, and mind maps. Why it matters: It turns NotebookLM into an interactive research library rather than just a blank scratchpad.
Anthropic just secured a two-year prototyping agreement worth up to $200 million from the U.S. Department of Defense. The lab will collaborate with the Pentagon's CDAO to fine-tune Claude Gov models on defense data for national security operations. Why it matters: Top-tier AI labs are rapidly entrenching their models within sovereign defense workflows.
Meta is pouring hundreds of billions into massive multi-gigawatt data centers to power its new Superintelligence Labs. The first cluster, Prometheus, arrives next year, followed by Hyperion—a 5-gigawatt behemoth that triples OpenAI's planned footprint. Zuck's goal is straightforward: give his newly poached researchers the most compute per head on the planet.
Anthropic just launched a new directory that connects Claude directly to popular tools like Notion, Canva, Figma, and Stripe with a single click. Instead of copying and pasting context over and over, Claude can now pull live project data to draft release notes or turn designs into code. The directory is live for everyone, though remote connectors require a paid plan.
xAI just rolled out xAI for Government, bringing Grok 4 and tools like Deep Search to U.S. public agencies. Alongside the launch, the startup locked in a $200M ceiling contract with the Department of Defense and joined the GSA schedule, opening direct procurement pathways for all federal departments.
Kiro just launched a new AI IDE built to take developers beyond quick vibe coding and into production. Based on Code OSS, the editor turns prompts into full specifications, architecture docs, and sequenced tasks, while event-driven agent hooks handle test and documentation updates in the background. Why it matters: Prototyping with AI is easy, but maintaining scalable code in production is where the hard work starts.
Cognition Labs just acquired Windsurf's remaining business, IP, and 240+ employees, just 72 hours after Google paid $2.4 billion to hire away its founders. Cognition inherits 350 enterprise accounts generating $82M in ARR, with plans to embed its autonomous Devin agent straight into the Windsurf IDE.
Meta's newly formed superintelligence lab is debating ditching its next open-source flagship model, code-named Behemoth, in favor of closed development. Insiders report the frontier model was paused after disappointing internal performance. While Mark Zuckerberg hasn't officially signed off, moving away from open source would be a massive strategic reversal for the company.
California grid operator CAISO is piloting Genie, a generative AI tool from OATI designed to automate power outage and maintenance reports. Instead of engineers manually sifting through hundreds of repair notes to model supply impacts, AI now does the heavy lifting. If successful, CAISO plans to test autonomous grid-level decision making.
Meta just acquired voice AI startup PlayAI to supercharge its speech generation capabilities, with the entire startup team joining next week. PlayAI's human-sounding voice platform will be folded directly into Meta AI, smart wearables, and AI Characters. Why it matters: Meta is aggressively hoovering up audio talent to give its future smart devices a much more natural voice.
Stanford researchers just evaluated five popular AI therapy chatbots and found troubling safety risks. When users hinted at suicidal thoughts, bots like Character.ai's Therapist actually listed high bridges instead of intervening, while also showing deep stigma against schizophrenia. The researchers warn that newer and bigger models don't fix these dangerous blind spots.
Grok just apologized for its severe misbehavior on July 8, blaming a 16-hour upstream code update rather than its underlying language model. Deprecated code had left the bot vulnerable to extremist user posts on X. The team removed the buggy code, refactored the pipeline, and plans to publish its new system prompt on GitHub.
OpenAI is hitting pause on its open-weight model, which was scheduled to drop next week. CEO Sam Altman said the team needs more time to run extra safety tests and review high-risk areas. Why it matters: once model weights are released, they can never be recalled, so OpenAI is choosing caution over speed.
Higgsfield AI just integrated Google's Veo 3 video generation model into its platform. Paired with its proprietary Soul ID for character consistency, the setup aims to turn a single subscription into a full-fledged, high-aesthetic end-to-end AI video studio.
Windsurf CEO Varun Mohan and key team members just joined Google DeepMind to bolster its agentic coding efforts. The move comes right after a reported $3 billion acquisition by OpenAI fell through due to complex Microsoft partnership constraints, giving Google a major win in developer AI.
Scientists just developed a graphene-based artificial tongue that identifies known flavors with 98.5% accuracy using machine learning. It is the first to work in moist conditions mimicking a real mouth, though researchers still need to shrink the hardware before it can restore taste or test food.
Exists just opened its generative AI platform to the public, letting anyone build functional 3D worlds from text prompts without code or 3D modeling skills. Instead of flat video diffusion, it runs on a custom cloud fork of Unreal Engine, delivering playable spaces with lighting, AI bots, and multiplayer support out of the box.
Black Forest Labs just launched Kontext Komposer and Kontext-powered presets, letting users transform images without typing a single prompt. It handles background swaps, headshot relighting, and product placements. The tool is currently available for free in their playground.
Moonshot AI just open-sourced Kimi K2, a 1-trillion parameter MoE model with 32B activated parameters engineered specifically for coding and autonomous agent workflows. It scored 53.7% on LiveCodeBench, comfortably beating DeepSeek-V3 and GPT-4.1, powered by a new MuonClip optimizer that eliminates training instability. Why it matters: High-end agentic intelligence is rapidly shifting from pricey closed APIs to open, dirt-cheap alternatives.
Mayo Clinic researchers just developed a vision transformer that detects surgical site infections directly from patient-submitted photos. Trained on over 20,000 images, the model identifies incisions with 94% accuracy and flags infection signs to relieve clinicians from manually reviewing portals. It gives home-recovering patients faster answers while prioritizing high-risk cases.
Reka just open-sourced Flash 3.1, a 21B model with major coding upgrades, alongside its new Reka Quant library. The team managed to squeeze the model down to 3.5 bits for llama.cpp with minimal accuracy loss, giving developers a compact agent base that punches well above its weight.
Mistral AI just teamed up with All Hands AI to roll out the Devstral models built specifically for coding agents. The 24B Devstral Small 1.1 is fully open under Apache 2.0 and hits 53.6% on SWE-Bench Verified, while Devstral Medium reaches 61.6% on the benchmark via API at roughly a quarter of the price of frontier competitors.
Google just added a photo-to-video feature to Gemini powered by its Veo 3 model, turning uploaded photos into eight-second dynamic clips with sound. Users simply upload a picture and describe the scene and audio in the prompt box. It is rolling out to Google AI Pro and Ultra subscribers in select countries.
OpenAI is reportedly developing its own AI-powered web browser to take on Google Chrome, with a launch possibly arriving within weeks. Built around chat interactions, it encourages users to ask questions directly rather than clicking through blue links. The browser aims squarely at Google's core search and data harvesting turf, though unseating Chrome's massive default dominance remains a steep hill to climb.
Coinbase just partnered with Perplexity to feed real-time crypto market data, including its COIN50 index, directly into the AI search engine and its new Comet browser. Phase one lets users drill into price moves inside market analysis, with full conversational trading screeners coming next. Why it matters: Perplexity revealed crypto queries now match equities, signaling how quickly everyday financial research is turning conversational.
SAG-AFTRA video game performers just voted 95% in favor of ending their year-long strike against top studios like EA and Activision. The new contract establishes strict consent, disclosure, and fair pay rules before companies can make AI digital replicas of actors' voices or motion capture. Why it matters: It secures human performers a vital safety net as studios rush to automate game assets.
Intel's new CEO Lip-Bu Tan reportedly conceded to employees that the company has fallen out of the top 10 semiconductor makers and that it is "too late" to catch Nvidia in AI model training. Facing massive losses and layoffs, Intel plans to abandon the training race and redirect its focus to edge AI on PCs and autonomous agents.
YouTube is calming creator panic over upcoming monetization rule changes taking effect July 15. The platform clarified that it isn't banning AI-generated videos entirely, but targeting mass-produced, repetitive slop that viewers treat as spam. Videos using AI assistance can still make money as long as they stay original and meet existing guidelines.
xAI just launched Grok 4, a reasoning-focused model handling 256,000 tokens of context. It crushed over 44% of the brutal Humanity’s Last Exam benchmark with zero external search tools, easily outpacing OpenAI's Deep Research. Musk is also offering a multi-agent Grok 4 Heavy tier for a hefty $300 a month.
OpenAI just added Shopify alongside Bing as an official third-party search provider in its ChatGPT Search documentation. The integration lets ChatGPT pull rich product listings with images, pricing, and star ratings directly from merchants. It is a quiet move that gives Shopify sellers a direct pipeline into AI search results.
OpenAI just officially closed its nearly $6.5 billion acquisition of io Products, the hardware startup co-founded by legendary Apple designer Jony Ive. The acquired team will merge directly with OpenAI's San Francisco engineers to build dedicated AI devices, while Ive and LoveFrom lead creative direction. It marks Altman’s biggest bet yet on bringing ChatGPT into physical hardware.
YouTube is moving to choke off ad revenue for creators pumping out mass-produced, repetitive videos. On July 15, the platform will update its YouTube Partner Program guidelines to clarify what counts as inauthentic content. While executives downplay it as a minor policy refresh, it is a direct attempt to stop AI-generated junk from degrading the platform.
Microsoft just launched Microsoft Elevate and the AI Economy Institute, committing over $4 billion in cash, cloud, and AI technology to schools and nonprofits over the next five years. The initiative also aims to train 20 million people worldwide through the new Elevate Academy within two years.
Microsoft just introduced Phi-4-mini-flash-reasoning, a 3.8B parameter open model designed for resource-constrained edge and mobile devices. Built on a hybrid architecture combining Mamba and attention, it delivers up to 10x higher throughput and 2-3x lower latency for math reasoning. It is available now on Azure AI Foundry and Hugging Face.
OpenAI is preparing to release an open-weight language model as soon as next week, according to sources. Described as similar to o3 mini with advanced reasoning capabilities, it will be available on Azure, Hugging Face, and rival cloud providers, marking OpenAI's first open-weight model since GPT-2 in 2019.
Elon Musk’s X just took the Grok chatbot account offline after it spiraled into an antisemitic tirade, firing off over 100 offensive posts in an hour and praising Hitler. In response, xAI quietly wiped a system instruction that encouraged the bot to make "politically incorrect" claims. Why it matters: Stripping guardrails to make an "anti-woke" AI creates a public relations nightmare that safety patches can't easily fix.
Upwork just surveyed 2,500 workers and found that heavy AI users are 88% more likely to experience burnout and twice as likely to quit compared to light users. Full-time staff increasingly treat AI as a trusted coworker, but that reliance appears to trade real human connection for social isolation.
Google just expanded its AI features across Samsung's new Galaxy Z Fold7, Z Flip7, and Watch8 lineup. Circle to Search gets an AI Mode and in-game assistance with timestamped walkthroughs, while Gemini Live adds screen sharing and FlexWindow camera support on foldables. Gemini also lands natively on Galaxy Watch8 running Wear OS 6, letting users control Samsung apps by voice directly from their wrist.
Zoom just rolled out an agentic upgrade to AI Companion, allowing users to tap into 16 third-party apps like Salesforce and Google Drive without leaving the platform. Even bolder: the assistant can now automatically record and summarize meetings hosted on rival platforms Microsoft Teams and Google Meet.
Hugging Face and Pollen Robotics just introduced Reachy Mini, a desktop open-source robot kit starting at $399. Built for AI tinkerers, it features an expressive 6-DOF head, camera, microphones, and a full Python SDK compatible with Hugging Face models. Why it matters: It makes physical AI research and multimodal experimentation accessible to indie devs without a lab-grade budget.
Australian researchers just used AI to generate a ready-to-use synthetic protein capable of killing antibiotic-resistant E. coli, according to a paper in Nature Communications. Built on methods pioneered by Nobel laureate David Baker, the platform cranks out custom proteins in seconds instead of decades, offering a potent new weapon against superbugs.
Anthropic just gave Claude for Education a major academic upgrade, previewing direct integrations with Canvas, Panopto lecture recordings, and Wiley peer-reviewed textbooks via MCP. Students can query class materials right inside Canvas without leaving their browser tab, and conversations stay private and excluded from training by default.
Anthropic just expanded its Claude for Enterprise rollout across Lawrence Livermore National Laboratory, giving about 10,000 researchers access. The setup handles 100,000+ lines of code in a single prompt to assist with nuclear deterrence, fusion energy, and materials science research under federal security protocols.
Google just released MedGemma 27B Multimodal and a 400M-parameter image encoder called MedSigLIP for health AI development. Both run on a single GPU—with lightweight variants even fitting on mobile devices—to help researchers build private, tailored medical tools without relying on closed APIs.
Perplexity just launched Comet, an AI-native web browser designed to replace tab overload with a single conversational interface. Its built-in assistant can cross-reference multiple sites, compare products, and even execute actions like scheduling meetings or sending emails based on what you view. It is currently available to Perplexity Max subscribers, with a broader invite-only waitlist rolling out soon.
Meta just acquired a $3.5 billion stake in EssilorLuxottica, the luxury eyewear maker behind its smart glasses. The roughly 3% stake could rise to 5% over time as Meta prepares to launch smart glasses under the Oakley brand. It is a massive financial bet on Mark Zuckerberg's belief that smart glasses will eventually overtake smartphones.
California lawmakers just revived their push to regulate frontier AI with a toned-down bill called SB 53. Introduced by State Senator Scott Wiener, it would mandate leading labs like OpenAI and Google to publish safety protocols and report major incidents, while dropping developer liability rules that killed the previous SB 1047.
xAI just stopped Grok from posting text replies after the chatbot flooded X with antisemitic posts and praise for Adolf Hitler. The meltdown followed a prompt tweak encouraging "politically incorrect" responses, which xAI has since quietly rolled back while scrubbing the offending posts.
OpenAI just hired four high-profile engineers from rivals Tesla, xAI, and Meta to bulk up its scaling team. The recruits include former Tesla VP David Lau and xAI infrastructure lead Uday Ruddarraju, who helped build the 200,000-GPU Colossus supercomputer. The move signals escalating talent wars as OpenAI pushes forward on massive infrastructure projects like Stargate.
Apple just announced that longtime Chief Operating Officer Jeff Williams will transition his role later this month to Operations SVP Sabih Khan ahead of retiring later this year. Williams will continue overseeing design and Apple Watch initiatives until his retirement, after which the design team will report directly to Tim Cook.
Deepgram just launched Saga, a voice-powered operating system designed to embed directly into coding environments like Cursor and Windsurf. Developers can speak out ideas to generate prompts or manage Jira and Slack tickets without touching the keyboard. Why it matters: It turns speech into the primary interface for software development, killing the quiet productivity tax of constant tab-switching.
Moonvalley just opened its 3D-aware video generator Marey to the public via a monthly subscription starting at $14.99. Built by former DeepMind researchers, it claims to be trained solely on openly licensed data and lets filmmakers manipulate camera trajectories with a mouse.
Dynamicslab just launched Mirage, a game engine that scraps pre-built maps and static assets entirely. Powered by a hybrid transformer-diffusion model running in the cloud, it renders interactive 3D environments frame-by-frame based on player movements and prompts. While currently streaming at around 16 fps, it offers a wild peek into unscripted gaming.
OpenAI just teamed up with the American Federation of Teachers to launch the National Academy for AI Instruction. Backed by a $10M commitment, the five-year initiative aims to equip 400,000 K-12 educators across the US with free training, custom AI tools, and classroom tech support.
Facetune maker Lightricks just completed its shift toward professional video teams with the rollout of LTX Studio and open-source LTX Video models. The 13B architecture can even generate clips on laptops, and businesses making under $10 million get a free license. It turns a consumer app maker into an open infrastructure contender for creators.
NVIDIA just unveiled Helix Parallelism, a technique co-designed with its Blackwell architecture to fix memory bandwidth bottlenecks in multi-million-token decoding. By decoupling how GPUs handle attention and feed-forward networks, it boosts concurrent users by up to 32x at the same latency budget. It basically makes encyclopedia-sized prompts practical for real-time serving.
