Elon Musk just led an unsolicited $97.4 billion bid to acquire OpenAI, aiming to force the ChatGPT maker back into an open-source, safety-focused nonprofit with help from xAI. Sam Altman immediately laughed off the hostile takeover on X, sarcastically offering to buy Twitter back for $9.74 billion instead.
Google just bundled NotebookLM Plus into its $19.99/month One AI Premium plan at no extra charge. Subscribers get five times more Audio Overviews, notebooks, and queries, plus custom chat responses and viewer analytics. It makes Google's AI bundle a lot more compelling for heavy researchers.
Google, OpenAI, Roblox, and Discord just launched ROOST, a non-profit providing free open-source AI tools to detect and report child exploitation material. Backed by over $27 million, the group plans to share vetted datasets and host moderation APIs. Why it matters: Amid intense regulatory pressure, tech giants are pooling resources to set shared safety standards.
Luma AI just rolled out image-to-video generation powered by Ray 2 inside Dream Machine. Users can drop in virtually any still picture—historical paintings, artifacts, personal photos, or internet memes—and animate them into motion. It is available right now.
Lyft is teaming up with Intel's Mobileye and Japanese conglomerate Marubeni to launch a robotaxi fleet in Dallas as soon as 2026. The ridehail platform is sticking to an asset-light playbook, letting Marubeni own and manage the cars while Mobileye supplies the self-driving tech, all in a bid to keep pace with Uber.
Apple engineers just built an expressive robotic desk lamp straight out of a Pixar film. Equipped with cameras, microphones, and tiny motors, the prototype responds to hand gestures, grooves to music, and even peeks out the window before answering weather questions.
French President Emmanuel Macron announced that France's AI sector secured 109 billion euros in private investment for the coming years. Backed by UAE funds, North American investors, and local tech firms like Mistral, the push aims to serve as Europe's heavyweight answer to the US Stargate initiative.
Anthropic just launched the Anthropic Economic Index to track AI's real-world impact on labor markets using anonymized Claude conversations. The data shows AI leans toward human augmentation over automation at 57% to 43%, with software engineering and writing seeing the heaviest use.
OpenAI just struck a content partnership with Nordic powerhouse Schibsted Media Group to pull articles from major titles like VG and Aftenposten directly into ChatGPT. The integration provides ChatGPT's 300 million users with timely Nordic news summaries featuring clear source links. Why it matters: OpenAI continues building out its verified media network to feed real-time responses while shielding itself from copyright backlash.
Researchers just introduced Animate Anyone 2, an upgraded image animation framework that finally solves character-environment interactions. Instead of producing floating figures pasted onto backgrounds, it extracts environmental cues and uses a dedicated object guider so animated characters can naturally touch, sit on, and interact with surrounding props. It is a huge leap toward realistic video generation with consistent physical affordance.
OpenAI just aired its first-ever Super Bowl ad, charting human progress from fire to ChatGPT to usher in what it calls the Intelligence Age. The company shared that over 300 million people now use ChatGPT—with nearly eight in 10 under 35—and that its creative team leaned on Sora to visually brainstorm the campaign.
Groq just opened its Developer Tier to everyone with no waitlist or prerequisites. Developers get higher rate limits, global billing support, and flex and batch processing options. It removes the scaling bottleneck for teams building real-time AI apps on fast LPU inference.
Researchers just introduced VFX Creator, an efficient framework that turns static images into animated cinematic effects. By pairing text prompts with mask controls and start-end timestamps, users can pinpoint exactly where and when an effect triggers. The system is built on a Video Diffusion Transformer and trained on the new Open-VFX dataset covering 15 effect categories.
OpenAI just shared three key economic observations on the road to AGI, noting that intelligence costs drop 10x every 12 months while socioeconomic value grows super-exponentially. The post also highlights that AI agents will soon act as junior virtual coworkers, making human agency more vital than ever.
Christie’s is hosting its first auction dedicated solely to AI-generated art, spotlighting works trained on artists' own inputs and a robot painting live. But the backlash was immediate: an open letter demanding the auction's cancellation collected over 1,800 signatures in less than 24 hours over copyright and uncredited training data.
MIT and Whitehead Institute researchers just introduced ProtGPS, a deep-learning model that predicts how proteins sort themselves within cellular condensates. Built on Meta's ESM framework, it complements AlphaFold's structural predictions by decoding spatial organization rules, paving the way for targeted drug discovery.
Andrew Ng just backed Google's decision to drop its seven-year pledge against building AI weapons. Speaking at a veterans event, the Google Brain cofounder argued American firms shouldn't refuse to help U.S. troops, stressing that competing with China is vital. The shift marks a broader pivot as tech giants seek lucrative military partnerships.
Perplexity just rolled out file and image uploads for free account holders, complete with a massive 1 million token context window. That means anyone can now dump an entire 400-page book or 1,000 pages of PDF financial reports into the search tool at once, easily trouncing ChatGPT's paid 128k context cap.
Safe Superintelligence, the AI startup founded by former OpenAI chief scientist Ilya Sutskever, is reportedly in talks to raise funding at a valuation of at least $20 billion. The pre-revenue startup was valued at $5 billion just last September, quadrupling its valuation as investor hype around top frontier AI researchers shows no signs of slowing down.
Krea AI just rolled out Krea Chat, an open beta tool powered by DeepSeek that brings image and video creation directly into a conversational interface. Users can generate visuals from text, borrow styles from uploaded references, or convert generated images into short video clips. It is free to try with a registered account, though compute limits apply.
OpenAI just updated the chain of thought for o3-mini, revealing far more reasoning steps to both free and paid ChatGPT users. While it still holds back raw internal monologues to prevent distillation, a new post-processing layer cleans up unsafe output and translates the thought process into non-English native languages.
OpenAI just increased the cross-chat memory limit in ChatGPT by 25% for Plus, Pro, and Team subscribers. The upgrade gives the chatbot more room to store ongoing preferences, details, and context across workflows without forgetting earlier notes. It is already live for paid accounts.
Pika just launched Pikaswaps, a new video editing feature that lets creators swap out specific objects, clothing, and backgrounds inside video clips. Users can brush over a target area and replace it using text prompts or reference images, making complex VFX and green screen composites as easy as prompting.
GitHub just introduced agent mode for Copilot in VS Code alongside the general availability of Copilot Edits. The agent can infer missing subtasks, propose terminal commands, and fix runtime errors on its own without manual copy-pasting. GitHub also added Gemini 2.0 Flash to the model picker and previewed Project Padawan, an autonomous agent built to resolve issues and open pull requests directly.
Pika just launched Pikadditions, a new video editing feature that lets you insert any AI-generated person or object into existing footage. Whether it is your own recording or a classic clip, the tool blends elements straight in, and new signups get 15 free generations right out of the gate.
AI researcher Steven Hoi just joined Alibaba to lead its AI-to-Consumer business group. He plans to focus on building foundation models and AI agent applications for consumer markets over the coming decade. It shows Alibaba is getting serious about turning frontier model tech into everyday consumer software.
Mistral just rolled out dedicated iOS and Android apps for its le Chat assistant alongside a major platform upgrade. It packs web search, document processing, Flux Ultra image generation, and a new Flash Answers mode that replies over 10x faster than ChatGPT.
OpenAI is reportedly set to air its first television commercial during Sunday’s Super Bowl, according to The Wall Street Journal. Coming shortly after the December hire of its first CMO, Kate Rouch, the spot marks the ChatGPT maker’s biggest advertising push yet. Why it matters: As tech giants race for mainstream users, OpenAI is trading pure word-of-mouth for prime-time marketing.
Lyft just partnered with Anthropic to deploy Claude for customer support, cutting resolution times for driver inquiries by 87%. The bot handles frequent questions before routing trickier issues to human staff. Notably, Lyft also revealed that generative AI now writes one in four lines of its engineering code.
Topaz Labs just introduced Project Starlight, calling it the first-ever diffusion model designed specifically for video restoration. The tool reconstructs old, low-resolution clips into crisp high-definition footage, which the company bills as its biggest leap since Video AI debuted. Early access is currently opening up.
Kyutai just introduced Hibiki, a simultaneous speech-to-speech translation model currently handling French to English. It delivers real-time voice and text translations while keeping the speaker's original timbre and pacing naturally. Evaluations show it rivals professional interpreters in naturalness and accuracy.
Google just opened up its flagship Imagen 3 image generation model to developers via the Gemini API. Priced at $0.03 per image, it delivers high-fidelity visuals across styles while embedding SynthID watermarks. Paid API users get access first, with a free-tier rollout coming soon.
OpenAI co-founder John Schulman just left rival lab Anthropic after only six months on the job. Schulman, who originally joined to focus on hands-on AI alignment research, announced on X that he is departing to pursue another "extremely compelling" opportunity.
AI researchers at Stanford and the University of Washington just created an open-source reasoning model called s1 for under $50 in cloud compute. By distilling thinking steps from Google's Gemini into a Qwen base model using 1,000 curated questions, s1 matches frontier reasoning models like OpenAI's o1 on math and coding tests.
OpenAI just opened up ChatGPT Search to the public without requiring an account or login. Anyone visiting the site can now type a query to get real-time news, sports scores, and stock quotes directly from the web, though switching to the reasoning mode still requires signing in.
Andrej Karpathy just released a massive 3.5-hour video breaking down the entire LLM training pipeline for everyday users. The tutorial walks through pretraining, supervised finetuning, and reinforcement learning, covering everything from tokenization to model hallucinations. It is basically the ultimate free, intuitive masterclass on modern AI.
Google just made Gemini 2.0 Flash generally available to everyone and introduced an experimental Gemini 2.0 Pro packing a massive 2M-token context window for heavy coding and reasoning tasks. The lineup also adds Flash-Lite, offering 1.5 Flash performance at cut-rate pricing. Gemini Advanced users can already pick the Pro model from the app switcher.
OpenAI just completed the rollout of its Deep Research feature to all Pro subscribers across Europe. The tool turns the web into a personal research assistant, leveraging autonomous agent capabilities to tackle complex research tasks. European Pro users can now run in-depth reports directly in ChatGPT.
Amazon is set to preview a major generative AI overhaul of Alexa at a February 26 event in New York. The revamped assistant can handle multi-turn conversations and act as an autonomous agent, partly powered by Anthropic's models. Why it matters: Turning Alexa from a glorified kitchen timer into an active assistant is Amazon's best shot at monetizing half a billion connected devices.
South Korea’s ETRI just developed an AI-powered sensor that distinguishes real fire smoke from harmless cooking fumes and dust. By measuring how light scatters across different wavelengths, the AI identifies non-fire particles to prevent nuisance alarms. The team plans to deploy it in aspirating smoke detectors first.
OpenAI just struck its biggest ChatGPT rollout to date, partnering with the California State University system to deploy ChatGPT Edu to over 500,000 students and staff across 23 campuses. The program includes custom course bots and free certification training to build an AI-ready workforce.
Researchers just introduced VideoJAM, a framework designed to tackle AI video's persistent struggle with real-world physics and motion coherence. By training models on joint appearance-motion representations and using internal motion cues to steer generation during inference, it plugs into existing architectures and outperforms proprietary baselines like Sora on complex motion.
Adobe just launched Contract AI for its Acrobat AI Assistant, an add-on feature that digests complex legal documents in seconds. It automatically recognizes contract PDFs, translates confusing jargon into plain English, and highlights version differences. Why it matters: It turns Acrobat into an affordable practical legal copilot for small businesses and individuals overwhelmed by fine print.
Virgin Media O2 built an AI bot named Daisy to troll phone scammers. Posing as a chatty 78-year-old grandmother who rambles about knitting and scone recipes, the bot kept fraudsters on the line for up to 40 minutes at a time to keep them away from real victims.
Researchers just introduced ASAP, a two-stage framework that bridges the sim-to-real physics gap for humanoid robots. By pairing simulated pre-training with a real-world residual action model, the team demonstrated agile whole-body skills on a Unitree G1 robot with significantly reduced tracking errors. Why it matters: It offers a practical path to dynamic robot agility without manual parameter tuning.
OpenAI just rolled out Deep Research to all ChatGPT Pro subscribers across iOS, Android, macOS, and Windows apps. The tool autonomously browses the web and synthesizes deep findings into structured reports, now accessible straight from your pocket or desktop.
LinkedIn is facing a proposed class-action lawsuit accusing the platform of unlawfully sharing Premium users' private InMail messages with third parties to train AI models. The suit claims LinkedIn quietly amended its privacy terms to cover its tracks and seeks $1,000 per user. LinkedIn denies using member messages for model training.
Researchers just introduced MatAnyone, a memory-based video matting framework tailored for complex scenes. By simply providing a segmentation mask on the first frame, the model leverages region-adaptive memory propagation to maintain semantic tracking while preserving delicate alpha boundaries. It aims to make Hollywood-level object separation accessible with just a few clicks.
President Trump just signed an executive order re-establishing the President's Council of Advisors on Science and Technology (PCAST), revoking the 2021 Biden-era framework. The revamped 24-member council will be co-chaired by the newly appointed Special Advisor for AI & Crypto alongside the top science advisor. Why it matters: The move officially elevates AI and crypto to the core of White House national security and tech policy.
Anthropic just rolled out Citations for the Claude API, allowing Claude 3.5 Sonnet and Haiku to trace claims back to exact sentences in source documents. It boosts recall accuracy by up to 15%, removes the need for brittle prompt engineering, and developers aren't charged for the quoted output tokens.
OpenAI CEO Sam Altman just confirmed that the company's upcoming lightweight reasoning model, o3-mini, will be available to all ChatGPT users, including the free tier. Paid Plus subscribers will receive generous usage allocations. The model focuses on faster and cheaper math and coding reasoning, though a specific launch date has not been set.
Perplexity just rolled out Perplexity Assistant for Android users, turning its search engine into an action-taking agent. It can hail rides, reserve tables, and read your screen or camera feed across apps in 15 languages. While completely free to use right now, CEO Aravind Srinivas admitted some complex multi-app actions might still be hit-or-miss.
Freepik just integrated Google's Imagen 3 into its AI Suite, touting its lighting and fine detail. Users can claim two free image generations within the first 24 hours of launch.
KREA AI just launched custom model training, allowing users to build personalized visual styles from as few as three reference images. Powered by io.net's decentralized GPU network, training takes only a few minutes, and the resulting style plugs directly into KREA's real-time generation canvas. The training feature is available on paid plans starting at $10 a month.
OpenAI just unveiled Operator, a research preview of an AI agent that controls a web browser to complete real-world tasks like filling out forms and ordering groceries. Powered by a new Computer-Using Agent (CUA) model, it navigates interfaces like a human while asking users to take over during logins and payments. It is currently rolling out to Pro users in the U.S.
OpenAI just launched a research preview of Operator, a general-purpose AI agent that takes control of a browser to handle travel booking, reservations, and online shopping. Powered by a Computer-Using Agent model, it clicks and fills out forms just like a human, though sensitive actions like payments still require manual user approval.
Yale School of Medicine researchers just built an AI tool that predicts future heart failure risk using standard 12-lead ECG images. The model was successfully validated across patient populations in the US, UK, and Brazil, giving doctors a low-cost screening method well before overt symptoms show up.
Elon Musk just questioned the massive Stargate AI project, claiming SoftBank and OpenAI don't actually have the money and SoftBank has secured well under $10B. OpenAI CEO Sam Altman fired back, saying what is great for the country isn't always optimal for Musk's companies.
Google has reportedly invested another $1 billion into Anthropic, pushing the tech giant's total stake to roughly $3 billion. The fresh cash arrives as Anthropic eyes a massive $60 billion valuation and preps an ambitious 2025 roadmap, which includes voice chat for Claude and a PC-controlling "Virtual Collaborator" agent.
Microsoft just launched an AI for Good Grants initiative in Washington state to celebrate its 50th anniversary. Local nonprofits, startups, and academic groups can apply to receive Azure compute resources, direct collaboration with Microsoft data scientists, and hands-on mentorship to tackle local challenges in healthcare, climate, and education.
Google is rolling out head and facial expression tracking to more Chromebooks, letting users navigate hands-free without a mouse. Built for users with motor impairments and based on its Project Gameface tool, it recommends at least 8GB of RAM. The company also unveiled new real-time classroom screen-sharing controls and teased over 20 upcoming laptops for 2025.
Adobe just rolled out major beta updates across its video suite ahead of the Sundance Film Festival. Premiere Pro gains on-device, AI-powered natural language footage search and 17-language caption translation, while After Effects overhauls its caching system to let older hardware preview complex comps smoothly. Finding needle-in-a-haystack B-roll just got a lot faster.
Samsung just unveiled the Galaxy S25 lineup at Unpacked 2025, turning the hardware into a vehicle for a new integrated AI platform powered by Gemini. A long-press of the power button lets the assistant run tasks across Samsung and Google apps, while adding features like video audio erasing. Why it matters: Phone AI is finally moving past chatbots to actually doing multi-step chores across your phone.
Spline just introduced Spell, a new diffusion-based AI model capable of generating full 3D environments from a single image in minutes. Outputs can be exported as Gaussian splats, meshes, or video while maintaining realistic multi-view physics and reflections without awkward morphing.
Google just updated its Gemini 2.0 Flash Thinking model, hitting 73.3% on math benchmark AIME and 74.2% on GPQA Diamond. The latest release introduces code execution, a 1M token context window, and reduces contradictions between reasoning steps and final answers. Why it matters: Google is moving fast on reasoning models, combining long-time planning research with frontier LLMs.
Microsoft just announced an evolution of its strategic partnership with OpenAI, locking in core terms through 2030 and teaming up on the Stargate supercomputer project. While Microsoft keeps Copilot IP rights and Azure API exclusivity, it loosened its capacity lock to a right of first refusal, clearing OpenAI to build additional training compute.
Anthropic is prepping a two-way voice mode and memory features for Claude to remember past conversations, CEO Dario Amodei just revealed at Davos. Smarter models are also slated for the coming months, though Amodei admits a recent demand surge has completely overwhelmed their compute capacity as the company seeks another $2B in funding.
X Games just partnered with Google Cloud to test experimental AI judging at the Aspen 2025 Superpipe competition. Powered by Vertex AI, the system analyzes video footage to calculate airtime and execution difficulty while generating multi-language commentary. Human judges still decide the podium, but live fans will get to see how the AI scores stack up against human eyes.
OpenAI, SoftBank, Oracle, and MGX just launched The Stargate Project, an ambitious venture planning to pour $500 billion into U.S. AI infrastructure over the next four years. With $100 billion deployed immediately starting in Texas, SoftBank handles the money while OpenAI runs operations alongside tech partners like NVIDIA and Microsoft.
ByteDance just launched Trae, a new AI code editor targeting overseas developers. Built on Microsoft software, it lets programmers generate project-level code using OpenAI's GPT-4o or Anthropic's Claude 3.5 Sonnet—and all AI features are currently free. The app is live on macOS, with a Windows version on the way.
Perplexity just rolled out its Sonar API, letting developers bake real-time web search with citations straight into their apps. The service comes in two tiers—standard Sonar and 200k-token Sonar Pro—with base pricing starting at $5 per 1,000 searches. Why it matters: Perplexity is productizing its core consumer moat into an API, taking direct aim at Google and OpenAI.
Meta is preparing a major wearables push for 2025, working on sports-focused Oakley smart glasses and a high-end pair with a built-in lens display dubbed Hypernova. The company is also exploring camera-equipped earbuds and smartwatches, aiming to weave its AI into more everyday hardware.
Tencent just open-sourced Hunyuan3D 2.0, a two-stage foundation system for generating detailed, textured 3D meshes from single images. By decoupling geometry creation via a DiT shape generator from texture baking with a dedicated diffusion model, it can also texture handcrafted meshes, beating previous open and closed-source baselines.
Mistral cofounder and CEO Arthur Mensch just shut down buyout rumors at Davos, stating the European AI champion is firmly "not for sale" and plans to pursue an IPO instead. Amid industry chatter about startup consolidation, the company is also opening a Singapore office and continuing its aggressive expansion across the US and Europe.
President Donald Trump just rescinded Biden's 2023 executive order on AI, immediately dismantling federal mandates that required developers to share safety test results before releasing advanced models. The new administration claims those guardrails handcuffed American innovation, and it has tasked AI adviser David Sacks with delivering a replacement action plan within 180 days.
Enterprise venture capitalists are moving their checkbooks past foundation models and into specialized applications. A TechCrunch survey of 20 enterprise investors reveals that the smart money in 2025 is hunting for task-specific agentic workflows, resilience tools, and AI infrastructure, while steering clear of single-feature startups.
DeepSeek just launched DeepSeek-R1, an open-weight reasoning model that matches OpenAI-o1's performance on math, code, and reasoning. It arrives under a permissive MIT license alongside six distilled smaller models, with chat and API access live immediately. Basically, developers can now freely commercialize and distill frontier-level reasoning outputs without restrictions.
Epoch AI just drew sharp criticism after revealing OpenAI funded its math benchmark FrontierMath while keeping contributors in the dark. OpenAI had early access to problems before demoing its flagship o3 model, sparking serious concerns over conflicts of interest and benchmark integrity.
OpenAI and other leading AI labs are preparing to unveil Ph.D.-level "super-agents" capable of independently tackling messy, multilayered human jobs, according to Axios. OpenAI CEO Sam Altman has even scheduled a closed-door briefing with U.S. government officials on Jan. 30. Why it matters: If these autonomous agents can overcome nagging hallucination issues, generative AI moves from an assistant to a real replacement for mid-level knowledge workers.
