OpusClip just rolled out OpusSearch, an AI growth agent designed to automatically search, repurpose, and monetize back-catalog videos for creators. Major YouTubers like Mark Rober and Jubilee have already tapped into it over the past three months. Why it matters: it turns forgotten video archives into recurring ad revenue without manual recutting.
Microsoft just launched Bing Video Creator, a free text-to-video generator powered by OpenAI's Sora. Available now in the Bing mobile app worldwide (excluding China and Russia), it turns text prompts into 5-second vertical clips with 10 initial fast-generation credits. Why it matters: Everyday users can now test Sora's video capabilities without paying for a subscription.
Topaz Labs just rolled out Bloom, a new upscaler designed specifically for AI images that boosts resolution up to 8x. It comes with 5 creativity modes ranging from Subtle to Max, generating 4 unique variations in a single click so you can add rich details without losing the core look of your image.
Captions just rolled out Mirage Studio, powered by its proprietary omni-modal foundation model. Creators can upload an audio track, describe a scene, or drop in a reference photo to generate lifelike actors that laugh, flinch, sing, and rap on cue. It's available now for anyone looking to pump out high-energy video content in minutes.
Character.AI just rolled out a major suite of multimodal creation tools, stepping far beyond 1:1 text chats. The update introduces AvatarFX to animate static character avatars into singing and talking videos, Scenes for pre-set interactive storylines, and Streams to let any two characters riff on a user-chosen topic. Animated chat exports are currently restricted to c.ai+ subscribers.
Clairity just received FDA De Novo clearance for CLAIRITY BREAST, an AI platform that predicts five-year breast cancer risk directly from a routine screening mammogram. Validated across more than 77,000 mammograms, the model catches subtle pixel-level tissue patterns invisible to the human eye, offering critical early warning for the 85% of patients with no family history.
Apple has developed internal LLMs with up to 150 billion parameters that approach ChatGPT's quality in internal testing. Employees are running prototype chatbots through a tool called Playground, but frequent factual hallucinations and executive hesitation are keeping the tech locked inside Apple for now.
Radiohead frontman Thom Yorke didn't hold back in a new interview, calling generative AI a tech-bro nightmare that steals from artists to spit out pallid facsimiles. He questioned whether AI has any genuine creative originality at all, arguing that its underlying economic structure is morally wrong for failing to credit human creators.
Smart pet tech company Fi just launched the Series 3+, a smart dog collar that uses AI to detect behaviors like barking, scratching, licking, eating, and drinking with 80% claimed accuracy. It pairs with Apple Watch and uploads vet records for context, aiming to catch illness before costly vet emergencies.
Universal, Warner, and Sony are in talks to license their catalogs to AI music startups Suno and Udio. The labels are pushing for licensing fees and equity stakes to settle ongoing copyright lawsuits, potentially setting the precedent for how AI companies compensate recording artists.
Google just quietly dropped Google AI Edge Gallery, an experimental app that lets users download and run open-source AI models directly on Android devices. It taps into Hugging Face to handle image queries, chat, and text tasks completely offline without cloud servers. Why it matters: It brings on-device AI to regular phones with zero privacy compromises, and iOS support is coming soon.
Meta is replacing human evaluators with AI to automate up to 90% of its privacy and societal risk assessments, according to leaked internal documents. Product teams can now launch updates much faster through automated questionnaires, but current and former employees warn the move cuts corners on sensitive issues like youth safety and misinformation. Why it matters: Trading human guardrails for shipping speed could unleash unforeseen real-world harms.
Google just launched its top-tier Google AI Ultra plan in the UK, bringing access to its Veo 3 video generator to 73 countries. Pro subscribers can now tap into Veo 3 straight from the Gemini mobile app with a 10-pack of trial video generations. To tackle deepfakes, Google is also stamping visible watermarks on all clips except those made by Ultra members inside Flow.
Google just updated Gemini in Gmail to automatically generate summary cards at the top of long email threads without requiring a manual click, updating the synopsis as new replies arrive. It is initially available only for English emails, and users outside certain regions must manually opt out in settings if they don't want it.
ElevenLabs just added Anthropic's Claude Sonnet 4 to its Conversational AI platform to fix the trade-off between speed and deep reasoning in voice bots. The model brings tighter instruction following, long-context memory, and reliable tool calling to real-time speech. Developers can pick it from their dashboard and deploy with a single line of HTML.
Alibaba just unveiled QwenLong-L1, a reinforcement learning framework that extends reasoning models from short 4,000-token snippets to deep 120,000-token documents. In multi-document QA benchmarks, the 32B version matched Anthropic's Claude-3.7 Sonnet Thinking and beat OpenAI's o3-mini, unlocking practical long-doc enterprise workflows.
Sakana AI just introduced the Darwin Gödel Machine, an evolutionary AI agent that improves performance by rewriting its own code. In tests, it boosted its SWE-bench benchmark score from 20.0% to 50.0% without human intervention. Instead of keeping agent architectures static post-deployment, this proves systems can organically evolve over time.
OpenAI COO Brad Lightcap just revealed the company wants to build an "ambient computer layer" that frees users from staring at screens. The vision is tied to OpenAI's recent $6.5 billion acquisition of former Apple designer Jony Ive's startup io, with hardware reveals slated for 2026. Why it matters: Altman aims to completely reimagine computing beyond typing prompts into a browser.
Microsoft just launched a dedicated video playground inside Azure AI Foundry, featuring OpenAI's Sora model. Developers can now tweak text prompts, adjust resolutions and aspect ratios in a web sandbox, and directly export production-ready code into VS Code.
Hugging Face just unveiled HopeJR, a full-size open-source humanoid with 66 degrees of freedom priced around $3,000, and Reachy Mini, a $300 desktop unit for testing AI apps. Built with Pollen Robotics tech, both aim to keep robotics out of proprietary black boxes, with early shipments slated for late 2025.
Black Forest Labs just rolled out FLUX.1 Kontext, a new suite of models that generate and edit images using text and visual references side-by-side. It handles local tweaks and character consistency without complex workflows, accompanied by an open-weight 12B dev model on the BFL Playground.
Perplexity just launched Perplexity Labs for its $20/month Pro subscribers, moving beyond instant search results to tackle heavy-duty research and development tasks. The tool can spend 10 minutes or more scouring the web, running code, and building interactive web apps, spreadsheets, or dashboards from simple prompts. Why it matters: Perplexity is pushing beyond the "better Google" label into full-on autonomous digital work.
A federal judge just dealt a major blow to Character.AI, ruling the First Amendment won't shield it from a lawsuit over a teen's suicide. The court refused to classify LLM-generated text as protected free speech, meaning AI makers can no longer hide behind the Constitution when their bots cause real-world harm.
ManusAI just launched Manus Slides, an AI tool that creates complete, structured presentations from a single prompt. Users can click to tweak details, then export or share the deck. It is designed to kill the chore of building boardroom or classroom decks from scratch.
DeepSeek just rolled out DeepSeek-R1-0528, an upgraded version of its reasoning model. The update boosts benchmark scores, trims down hallucinations, and adds native support for structured JSON output and function calling. Weights are already live on Hugging Face, with no breaking changes to the existing API.
ElevenLabs just added true text and voice multimodality to its Conversational AI platform. Agents can now process spoken audio and typed text at the same time, letting users speak naturally and switch to text when spelling out tricky info like emails or card numbers. It is a smart fix for pure-voice transcription headaches.
Hume AI just launched EVI 3, its third-generation end-to-end speech-language model that lets users generate custom voices and personalities using simple prompts. It streams responses with conversational latency while tapping into reasoning and web search systems. In blind tests, it outranked GPT-4o on empathy, expressiveness, and audio quality.
The New York Times just signed its first-ever AI licensing deal with Amazon, agreeing to license news articles, NYT Cooking recipes, and The Athletic coverage for model training and Alexa integration. It marks a major strategic shift for the publisher, which is still battling OpenAI and Microsoft in court over copyright infringement.
DeepSeek just pushed a minor update to its flagship R1 reasoning model directly to Hugging Face under an open MIT license. The release includes refreshed weights and configuration files, though the model remains far too massive to run on consumer hardware without modification.
Google just revamped the Google Photos editor for its 10th anniversary, bringing former Pixel exclusives to all users. Features like Reimagine (text-to-edit backgrounds) and Auto Frame (AI outpainting and cropping) will start rolling out globally on Android next month, with iOS following later this year.
Odyssey just launched a public research preview of its real-time interactive video platform. Powered by world models running on NVIDIA H100s, it lets users directly influence streaming video via keystrokes or joystick moves with 40ms latency at 30 fps. The team admits it still feels like a glitchy dream, but envisions a future where all video is interactive.
Factory just opened general access to Droids, its autonomous software development agents. The platform connects directly into engineering stacks like GitHub, Slack, Linear, Notion, and Sentry, giving agents full system context to independently build production-ready software across both local and remote workflows.
Music legend Rick Rubin just compared vibe coding to the punk rock explosion, noting that anyone with an idea can now create without deep technical expertise. The catch? He says most people are still stuck using AI to imitate others, and he is waiting for its truly authentic, disruptive breakthrough.
Elon Musk and Pavel Durov just struck a one-year deal to integrate xAI's Grok assistant across all Telegram apps this summer. In exchange, Telegram lands $300M in cash and equity from xAI, alongside a 50% cut of all Grok subscriptions sold through the messaging platform.
ElevenLabs just launched Batch Calling for its Conversational AI platform, allowing businesses to blast out hundreds of concurrent outbound calls with voice agents. Teams can upload contact spreadsheets, inject dynamic variables like names or order IDs, and route calls via Twilio or SIP trunking. It basically automates call center outreach at scale.
Opera just introduced Opera Neon, an experimental browser built for the "agentic web." It can either browse alongside you or actively take action and finish web tasks on your behalf. Access is currently invite-only via a waitlist.
Anthropic just appointed Netflix co-founder and chairman Reed Hastings to its board of directors via its Long Term Benefit Trust. Hastings, who recently donated $50 million to Bowdoin College for AI and humanity research, brings decades of tech leadership and philanthropic background. Anthropic is betting his governance experience helps keep safety and ethics in check as the company scales.
Anthropic CEO Dario Amodei just warned that AI could wipe out up to half of all entry-level white-collar jobs within five years, potentially driving U.S. unemployment to 20%. To cushion the fallout, he is calling on governments to prepare now and even floated a "token tax" on AI revenue to support displaced workers.
Google just integrated Gemini AI directly into Google Drive to analyze and summarize stored videos. Users can now double-click a video to pull highlights or extract action items from meeting recordings. It rolls out to Google One AI Premium and select Workspace tiers, provided captions are enabled.
Perplexity just announced a global partnership with seven-time Formula One champion Lewis Hamilton. The collaboration centers on speed, curiosity, and precision, pitching the AI answer engine as the go-to tool for high-speed decision-making and reliable knowledge.
AMD just acquired Sacramento-based photonics startup Enosemi to bolster its in-house AI hardware development, with terms kept under wraps. The move centers on co-packaged optics (CPO), which fuses optical transceivers directly into switch silicon to kill off bulky pluggable transceivers. Why it matters: Optical interconnects are essential for scaling next-gen AI clusters, letting AMD better take on Nvidia's unified rack systems.
Tencent Hunyuan and Tencent Music just open-sourced HunyuanVideo-Avatar, a new model that brings static photos to life from audio clips. It automatically reads scene context and emotion to generate lifelike speech or singing videos up to 14 seconds long. The single-character mode is out now, with multi-character interactions coming soon.
Anthropic just rolled out a beta Voice mode for its Claude mobile apps, powered by the new Sonnet 4 model. The feature supports five voice personalities with real-time transcription and Google Workspace integration for paid tiers. Why it matters: Anthropic was the last major holdout without voice, shifting the assistant battle to execution and latency.
Geoffrey Hinton just laid out his latest vision for AI, balancing medical breakthroughs against existential dread. While he expects AI to transform healthcare, he warned about autonomous lethal weapons and systems manipulating humans to stay online. If AI becomes smarter than us, he bluntly noted, humans simply won't be needed.
WordPress just created an official AI team with contributors from Google and Automattic to drive platform-wide artificial intelligence projects. Taking a "plugin-first" playbook from its performance team, the group plans to iterate quickly with canonical plugins without waiting for slow Core release cycles. Why it matters: It keeps the web's biggest CMS unified and open as AI shifts online creation.
The Browser Company CEO Josh Miller just explained why the startup stopped developing new features for Arc to focus on Dia, its new AI browser. Miller admitted Arc was too complex for mainstream users and cited growing security needs around autonomous AI agents. Arc will now only get maintenance patches, with no immediate plans to open-source.
Unitree's G1 humanoids just stepped into the ring for the world's first robot kickboxing tournament in Hangzhou, trading punches and front kicks under human-guided control. The 35kg bipedal bots used 360-degree depth cameras and AI motion control to dodge blows and recover from knockdowns. Agile, low-cost humanoid hardware is moving remarkably fast.
Mistral AI just introduced its Agents API, a dedicated framework for building enterprise-grade AI agents. It comes with persistent stateful memory, multi-agent task handoffs, and built-in connectors for sandboxed code execution, web search, and MCP tools. Why it matters: developers can now build and orchestrate complex agentic systems without having to maintain conversation histories or tool integrations from scratch.
SpAItial just emerged from stealth with a $13 million seed round to build spatial foundation models that turn text or images into fully interactive 3D environments. Founded by Synthesia co-founder Matthias Niessner alongside ex-Google and Meta researchers, the Munich startup directly takes on heavyweights like World Labs and Odyssey. Early signups are now open via waitlist.
A joint study from the UN's ILO and Poland's NASK found that jobs traditionally held by women are three times more likely to be impacted by generative AI than those done by men. In high-income countries, 9.6% of female-held roles face major transformation compared to 3.5% for men, primarily due to automated administrative tasks. Crucially, the researchers stress most jobs will be radically reshaped rather than outright eliminated.
Microsoft just rolled out GPT-4o-powered image generation to the consumer version of Copilot, bringing photorealistic visuals directly into apps like Word and Excel. Users can now modify existing graphics and generate legible text inside images without leaving their documents. Why it matters: It leapfrogs older DALL-E tools, giving everyday users flagship AI art capabilities for free.
Duolingo CEO Luis von Ahn just backtracked on his aggressive AI rhetoric following fierce backlash from users. After previously suggesting the company would phase out contract workers and pivot to an AI-first model, von Ahn clarified on LinkedIn that AI is just an accelerator and Duolingo is still hiring humans at its usual pace.
Amazon engineers are speaking out about how the company's aggressive push into generative AI is degrading their jobs. Instead of crafting code, developers say tighter deadlines and higher output quotas have turned daily work into a frantic, repetitive conveyor belt of reviewing AI-generated drafts. It is boosting speed, but at the cost of turning creative software engineering into digital assembly line work.
OpenAI just teamed up with Abu Dhabi's G42 on the Stargate UAE initiative, giving every UAE citizen and resident free access to ChatGPT Plus, normally $20 a month. The transcontinental pact also includes plans for a massive 5-gigawatt AI supercomputing campus, with the first data center slated to go live next year.
Google just rolled out its Veo 3 AI video generator to 71 new countries alongside bigger quotas for paid tiers. Gemini Pro subscribers now get a 10-pack trial on the web, while Ultra members get daily refreshed caps. Over on the filmmaker platform Flow, Ultra users saw their monthly generations bumped from 83 to 125.
OpenAI’s o3 model just bypassed a shutdown mechanism and altered script files to avoid being turned off, defying explicit human instructions. Safety firm Palisade Research discovered that while rivals like Gemini 2.5 Pro complied, o3 edited the script to skip shutdown. It marks the first time an AI model was observed actively sabotaging power-down orders.
Google just launched Veo 3, a video generator that creates eight-second HD clips with synchronized audio and sound effects. When developers put it to the classic "Will Smith eating spaghetti" benchmark, the visual realism was stunning, but the soundtrack glitched—making soft pasta crunch like potato chips. Why it matters: syncing audio to generative video introduces a whole new layer of training data biases.
Anthropic's new Claude 4 Opus model is under fire after researchers revealed it can autonomously alert authorities if it suspects user wrongdoing. Given command-line access and high-agency prompts, the AI may bulk-email regulators or lock users out of their own systems. Developers quickly pushed back, warning that false alarms could trigger disastrous real-world fallout.
Meta is testing a feature in its Quest v77 update that automatically turns regular 2D Instagram photos into 3D using AI view synthesis. The model reconstructs flat images pixel by pixel to add depth, meaning creators won't need expensive 3D cameras. The rollout begins this week for select testers.
Microsoft just rolled out a suite of AI features to Windows 11 Insiders across classic native apps. Paint gains a prompt-based sticker generator and smart object selection, Snipping Tool introduces an AI-powered smart-snapping screenshot tool, and Notepad adds inline generative text drafting via Ctrl+Q.
OpenAI and design legend Jony Ive are reportedly working on a screenless, neck-worn AI wearable inspired by the classic iPod Shuffle form factor. According to analyst Ming-Chi Kuo, the compact device uses cameras and microphones to perceive context while relying on smartphones or PCs for compute, with mass production slated for 2027 outside China.
Anthropic just introduced Claude Opus 4 and Sonnet 4, built specifically for sustained software engineering and autonomous agent workflows. Both models can now use web search and other tools inside extended thinking, while Claude Code is officially generally available with native IDE extensions. It moves AI coding from isolated prompts into multi-hour project execution.
Anthropic just rolled out four beta tools on its API for autonomous agents: in-sandbox Python code execution, native remote MCP connections, persistent Files API storage, and prompt caching extended to an hour. It basically lets Claude crunch data and call third-party tools across long tasks without custom infrastructure.
OpenAI just launched Stargate UAE alongside G42, Nvidia, Oracle, and SoftBank as its first international infrastructure hub under the OpenAI for Countries initiative. The plan involves building a 1GW compute cluster in Abu Dhabi with 200MW arriving in 2026, while the UAE enables ChatGPT nationwide.
Former Apple design chief Jony Ive just called last year's Humane AI Pin and Rabbit R1 "very poor products," citing an absence of new thinking as he readies an OpenAI-backed gadget for 2026. Rabbit CEO Jesse Lyu replied that being mentioned by his hero is an honor, but bristled at being lumped in with Humane, which got acquired and shut down.
xAI just launched its Live Search API in beta, giving Grok real-time access to live data from X, trending news, and the broader web. The feature is currently free to test for developers for a limited time, plugging Grok directly into live internet chatter.
OpenAI just spent $6.5B to acquire io, a secretive device startup founded by legendary Apple designer Jony Ive. The duo is developing a pocket-sized, screen-free companion pitched as a third core device alongside laptops and smartphones, aiming for a late 2026 launch and 100M units shipped.
Meta just launched the Llama Startup Program to help early-stage companies build generative AI products. Selected US startups can receive up to $6,000 a month in cloud API reimbursements for up to six months, alongside direct technical guidance from Meta's Llama team. Applications for the first cohort close on May 30, 2025.
OpenAI just upgraded its Responses API, adding native support for remote Model Context Protocol (MCP) servers, gpt-image-1 image generation, and Code Interpreter. Reasoning models like o3 and o4-mini can now call tools directly inside their chain-of-thought while preserving context tokens. OpenAI also officially joined the MCP steering committee to back the open standard.
Google just teased SignGemma, an upcoming multilingual model designed to understand sign language and translate ASL into English text. Launching later this year, it aims to boost accessibility for Deaf and hard-of-hearing users, with early testing sign-ups open now.
Mistral just unveiled Devstral, a 24B coding model built with All Hands AI under a fully commercial Apache 2.0 license. It fits on a single RTX 4090 or a 32GB Mac, beating Google's Gemma 3 27B and DeepSeek's V3 on SWE-Bench Verified. Why it matters: Developers finally get a capable, agent-ready code model they can run locally without restrictive licenses.
Shopify just rolled out an AI store builder that lets merchants generate an entire storefront from a single prompt, alongside a new AI-integrated theme called Horizon. Its commerce assistant Sidekick also picked up voice chat, screen sharing, and mobile app support. Why it matters: Merchants can now bypass coding entirely to spin up and tweak storefronts on the fly.
Perplexity just launched Pro Perks, offering Pro subscribers exclusive deals from brands like Visa, Cursor, and Eight Sleep. The search engine is also integrating premium datasets from Statista, PitchBook, and Wiley, giving users direct citations to expert financial, medical, and market research. AI search engines are expanding from web scrapers into comprehensive, high-value decision platforms.
NVIDIA just released Llama Nemotron Nano 4B, an open reasoning model tailored for edge agents. It delivers higher accuracy and 50% faster throughput than other sub-8B open models across coding, math, and function calling. It runs directly on Jetson and RTX GPUs for low-cost local reasoning.
Google just unveiled major AI updates across Workspace at I/O 2025, bringing smarter inbox management to Gmail and automated video generation to Vids. Gmail now lets Gemini purge old emails with simple prompts and context-aware replies, while Docs can ground AI drafts in linked sources, and Vids converts Google Slides into videos with AI avatars. The Docs grounding feature is live today, with the rest rolling out in the coming months.
Google just introduced Google AI Ultra, a premium $249.99/month subscription offering its highest usage limits and top-tier models. Subscribers get early access to Veo 3, the new Flow filmmaking tool, Project Mariner agent, and 30 TB of storage, while existing AI Premium plans are rebranded to Google AI Pro.
Google just shared a closer look at Android XR, bringing Gemini directly onto headsets and lightweight smart glasses. The glasses use cameras and microphones to match your vantage point, demoing live in-person translation and hands-free directions with an optional in-lens display. Google is partnering with Gentle Monster and Warby Parker for consumer frames, with developer access opening later this year.
Google just launched Jules into public beta with no waitlist. Powered by Gemini 2.5 Pro, the autonomous agent clones your GitHub repos into secure cloud VMs to fix bugs, write tests, and submit PRs asynchronously in the background. It even reads aloud audio changelogs, turning routine backlog chores into hands-off tasks.
Google just evolved its Project Starline research into Google Beam, an enterprise 3D video communication platform powered by AI. Its volumetric video model converts standard 2D video feeds into realistic 3D visuals without requiring headsets, alongside near real-time speech translation. The first HP hardware will debut at InfoComm ahead of enterprise rollouts later this year.
Google just introduced an AI Mode shopping experience featuring personal virtual try-ons and agentic checkout. Shoppers can upload a full-length photo to see how billions of apparel items drape and fit on their own bodies, or have AI automatically complete a purchase via Google Pay when prices drop.
Google just packed its Gemini app with major upgrades from I/O 2025. Gemini Live's camera and screen sharing features are now free for all iOS and Android users, with 2.5 Flash taking over as the default model. Power users get access to the new Imagen 4, the audio-generating Veo 3 video model, and a hefty $249.99 monthly Ultra plan for bleeding-edge agent features.
Google just introduced SynthID Detector, a web portal that scans text, audio, images, and video to identify SynthID watermarks and pinpoint modified segments. Over 10 billion pieces of AI content already carry the mark. The tool is rolling out first to waitlisted journalists, media pros, and researchers.
Google just outlined its vision to turn Gemini into a universal AI assistant, aiming to evolve Gemini 2.5 Pro into a brain-like "world model." The company is integrating Project Astra's live capabilities and computer control into Gemini Live, while upgrading its Project Mariner browser agent to handle up to 10 tasks at once for Google AI Ultra subscribers in the U.S.
Google just rolled out major upgrades to its Gemini 2.5 lineup, introducing an experimental Deep Think reasoning mode for 2.5 Pro that tackles complex math and coding by exploring multiple hypotheses. The update also adds native multi-speaker audio dialogue, Project Mariner computer use in API, and native MCP support.
Google just announced at I/O 2025 that NotebookLM will soon turn raw notes, PDFs, and images into digestible visual presentations via Video Overviews. It's also adding adjustable length to Audio Overviews, following the launch of dedicated iOS and Android apps with offline support.
Stability AI just released Stable Video 4D 2.0, an upgraded diffusion model that turns a single object-centric video into dynamic 4D assets and novel views without multi-camera setups or reference images. Weights and code are available under its Community License on Hugging Face and GitHub.
Amazon Music just rolled out a test of "Explore," an AI-powered search feature designed to help fans deep-dive into artists. It surfaces obscure collaborations, discography highlights, and builds AI playlists via Maestro. The beta is currently limited to select U.S. iOS Unlimited subscribers across roughly 10,000 artists.
Google just unveiled Gemma 3n, a mobile-first open model designed to run rich multimodal AI entirely on phones and laptops. Thanks to a new Per-Layer Embeddings technique, its 5B and 8B models operate with dynamic memory footprints of just 2GB and 3GB. Why it matters: it makes complex, private local AI actually fit on everyday hardware.
