A Chinese internet court just found an AI service liable for copyright infringement after generating images substantially similar to Ultraman, ordering 10,000 yuan in damages. The ruling, seen as a world first, puts AI training data and output liability front and center for developers everywhere.
OpenAI just asked a federal judge to dismiss key parts of The New York Times' copyright lawsuit, arguing the publisher used "deceptive prompts" and exploited bugs to force ChatGPT into spitting out verbatim text. The Times fired back, claiming it was merely hunting for evidence of unlicensed copying.
Automattic is reportedly finalizing deals with OpenAI and Midjourney to feed user posts from Tumblr and WordPress.com into their AI models. An internal scrape reportedly scooped up a decade of posts, including some private content by mistake, prompting Automattic to promise a quick opt-out toggle. Why it matters: Platforms are desperate to cash in on user data, even at the cost of alienating the creative communities that built them.
Pika just rolled out Lip Sync powered by ElevenLabs, allowing AI-generated characters' mouths to move in sync with spoken dialogue. Creators can type text prompts or upload their own audio tracks directly. For now, it is available in early access only to Pika Pro subscribers and invited Super Collaborators.
GitHub just made Copilot Enterprise generally available for $39 per user per month. The new tier hooks directly into your company's private codebase and institutional knowledge, letting developers ask questions on GitHub.com, draft pull request summaries, and browse the web via integrated Bing search. It turns AI from a simple code autocomplete tool into an org-wide engineering partner, while promising never to train models on your private repos.
Superhuman just launched Instant Reply, a feature that precomputes three complete draft replies for every incoming email. Unlike basic short suggestions, these are full responses tailored to match your personal tone, ready to send with just Tab and Enter. Early testers reportedly replied twice as fast.
Alibaba researchers just unveiled EMO, an expressive audio-driven portrait-video generation framework. Give it a single still image and an audio clip—talking or singing—and it animates the portrait with realistic facial expressions and head poses for any duration. It even keeps up with fast-paced rap and works across multiple languages, paintings, and 3D models.
Google DeepMind just unveiled Genie, an 11-billion-parameter foundation world model that conjures playable 2D platformers from a single image or sketch. Trained on over 200,000 hours of unsupervised gameplay videos, it automatically infers character controls and physics. Why it matters: It turns static art into interactive worlds and could serve as an endless simulator to train future autonomous AI agents.
Google's Demis Hassabis confirmed the company plans to restore Gemini's image generation of people in the next few weeks after pausing it over historical inaccuracies. Hassabis explained that well-intended diversity prompts were applied too bluntly across historical contexts, and fine-tuning is underway.
Microsoft just announced a multi-year partnership with French startup Mistral AI, making its new flagship model Mistral Large available first on Azure AI. Mistral gets Azure supercomputing power for model training, while Azure customers can run the model via MaaS alongside OpenAI's lineup. Why it matters: Microsoft is aggressively diversifying its cloud AI portfolio beyond OpenAI.
ElevenLabs just partnered with Perplexity to launch Discover Daily, a short-form daily podcast covering breakthroughs in innovation, science, and culture. The show merges Perplexity's real-time search engine with ElevenLabs' synthetic voices to deliver automated audio briefings on the go, available now across major podcast platforms.
Microsoft just added a new Generative Erase tool to the built-in Photos app, bringing pixel-style object removal straight to your PC. Even better, Microsoft is backporting the feature and background removal tools to Windows 10 and Arm64 PCs instead of forcing hardware upgrades. It is currently rolling out to Windows Insiders.
Suno just rolled out early access to its V3 Alpha model for Pro and Premier subscribers. The update bumps maximum clip lengths to two minutes, adds dedicated instrumental support, improves audio quality, and lets creators continue tracks from any point. Paying accounts also received 300 free credits to test the model.
OpenAI just updated the GPT Store with a 1-to-5 star rating system and private creator feedback. Builders can now also link their LinkedIn, X, and websites on expanded profile pages while tracking total conversation stats. The promised revenue-sharing model, however, is still nowhere to be seen.
The Justice Department just appointed Princeton professor Jonathan Mayer as its first chief AI officer and chief science advisor. Mayer will build an expert team to help federal law enforcement keep pace with rapidly evolving tech. Why it matters: In line with the Biden administration's executive order, his team will help establish crucial fairness guidelines for AI in sentencing, parole, and surveillance.
Google just deepened its partnership with Reddit, gaining direct access to Reddit's Data API to show fresher forum content in search and train its AI models. In return, Reddit will use Google Cloud's Vertex AI to overhaul its platform search. Why it matters: authentic human discussions have become the rarest commodity in the AI race, and Google just plugged straight into the web's largest conversation board.
Google just rolled out an experimental 'Help me write' feature in Chrome M122 for Mac and Windows. Powered by Gemini, the tool understands webpage context to draft or polish reviews, inquiries, and emails directly inside text fields. It is currently available to English users in the U.S.
AI marketing copilot Jasper just acquired image platform Clipdrop from Stability AI, closing the deal on February 20. The Paris-based team joins Jasper immediately to spearhead multimodal research. Enterprise clients get instant access via the Jasper API, while the standalone consumer app stays live.
Stability AI just announced Stable Diffusion 3 in early preview, calling it its most capable text-to-image model yet. The suite ranges from 800M to 8B parameters and combines diffusion transformer architecture with flow matching to solve messy spelling and multi-subject prompts. A waitlist is now open ahead of an open release.
ElevenLabs just announced it was selected for the 2024 Disney Accelerator program. The voice AI startup aims to tap into Disney's massive creative sandbox to break down language barriers. Why it matters: Landing a spot with Disney gives ElevenLabs a direct pipeline into Hollywood's biggest media powerhouse.
Google just shut off image generation of people in its Gemini chatbot after the tool churned out historically inaccurate and overcorrected images. The company admitted its diversity safeguards went too far, causing the model to misapply diversity to specific historical contexts and become overly cautious.
Tech giants are cutting jobs, and everyone is blaming AI — but experts say that's mostly a cover story. Oxford researchers and analysts point out there's little evidence of robots directly replacing workers at scale; instead, companies are just cutting standard costs or slashing other budgets to fund expensive AI pivots.
Match Group just inked an enterprise deal with OpenAI, buying over 1,000 licenses to help employees with coding, design, and analysis. In peak dating-app fashion, the announcement was written with ChatGPT and packed with cringey romance puns. It's all part of Match's $20M+ bet on AI this year.
Adobe just set up a new 50-person research group called CAVA, short for Co-Creation for Audio, Video, & Animation. Formed right after OpenAI dropped Sora, the team focuses squarely on generative AI for motion and sound. It signals Adobe is gearing up to defend its creative turf against frontier multimodal models.
Google DeepMind just created a new AI Safety and Alignment organization to tighten safeguards around Gemini and tackle future AGI risks. Led by UC Berkeley professor Anca Dragan, the U.S.-based unit will expand beyond London's existing alignment team to address current issues like misinformation and medical advice while researching superintelligent control. Why it matters: Google is scrambling to show regulators and enterprise buyers that its models can be trusted without flying off the rails.
Google just added its Gemini AI directly into Gmail, Docs, Sheets, Slides, and Meet for Google One AI Premium subscribers. For $19.99 a month, users can draft emails and build budgets without hopping between tabs or copy-pasting across apps. The English rollout spans over 150 countries, backed by a two-month free trial.
Google just released Gemma, a new lightweight family of open models built on the same research as Gemini. Available in 2B and 7B parameter sizes, it runs directly on laptops or desktops, beats rival open models at its scale, and comes with permissive commercial licensing. It gives local and open-source AI developers a massive new alternative to Llama.
Groq just went viral with an AI chip architecture that makes standard cloud GPUs look sluggish. Powered by custom Language Processing Units (LPUs), independent benchmarks clocked Groq hitting 247 tokens per second—over 13 times faster than Microsoft’s cloud baseline. Why it matters: this raw speed bypasses memory bottlenecks, finally making lag-free, real-time voice conversations with AI viable.
Will Smith just dropped a hilarious Instagram clip parodying the infamous 2023 AI video of him devouring pasta. Coming right after OpenAI debuted Sora, Smith's real-life recreation highlights how quickly AI video generation evolved from a surreal nightmare meme to uncanny realism.
Fresh off his OpenAI exit, Andrej Karpathy just released a two-hour deep dive on building a GPT tokenizer from scratch. The tutorial breaks down why LLMs stumble on basic tasks like string reversal and arithmetic, while explaining why YAML often beats JSON for saving precious tokens.
Microsoft just confirmed that OpenAI's Sora video generator will eventually make its way to Copilot. Advertising and web services CEO Mikhail Parakhin replied on X that integration is planned, though it will take time. There is no official timeline yet, but bringing text-to-video directly into Copilot promises to significantly expand its multimodal creative toolkit.
The AI crypto sector just added $10.21 billion in market value over 17 days, leaping 142.79% following OpenAI's reveal of text-to-video model Sora. Heavyweights like TAO jumped 145.38% alongside widespread triple-digit gains. Why it matters: Hype around generative AI can instantly reroute billions into speculative tokens, even before the model is publicly accessible.
Elon Musk's X is reportedly exploring a partnership with text-to-image generator Midjourney. If finalized, the deal could integrate Midjourney's visual capabilities directly into Grok, letting users generate and post images right from their feeds. While details remain sparse, the feature will likely be locked behind an X Premium subscription.
OpusClip just rolled out version 3.0, introducing genre-specific curation models for formats like Q&As and webinars. It can generate clips up to 15 minutes long, runs 300% faster on hour-long footage, and adds AI b-roll and standalone captioning tools.
ElevenLabs just announced that AI Sound Effects are coming soon to its platform. The feature allows creators to generate custom sound effects, ambient audio, and soundscapes from text prompts, expanding the company's toolset beyond voice and music.
Air Canada just lost a lawsuit after its AI customer service chatbot made up a fake refund policy for a grieving passenger. The airline tried to argue in court that the bot was a "separate legal entity" responsible for its own actions, but the tribunal flatly rejected it, ruling companies are fully liable for what their AI tells customers.
Reddit just inked a $60 million content licensing deal with an unnamed major tech company to train AI models, according to Bloomberg. Ahead of its planned $5 billion IPO in March, the platform is turning its vast forum discussions into cash, though the move risks reigniting fury among users already frustrated by recent API price hikes.
ElevenLabs just introduced new safeguards to prevent its AI voice tech from meddling in the 2024 elections. The company rolled out a "no-go voices" safeguard that blocks users from mimicking active political candidates, starting with US and UK races, and explicitly outlawed political robocalls and impersonation chatbots. Why it matters: With critical elections taking place worldwide, the pressure is on AI voice platforms to curb convincing deepfakes before they derail voter trust.
NBA commissioner Adam Silver just revealed NB-AI, a voice-driven assistant built to personalize live basketball broadcasts. In a demo with rookie Victor Wembanyama, Silver commanded the tool to remake a live play "as if it were a Spider-Man movie," instantly transforming players into comic-style cartoon characters and letting viewers buy sneakers off players' feet in real time.
Twenty leading tech companies just teamed up to curb deceptive AI in elections. At the Munich Security Conference, companies including Meta, Google, Microsoft, and OpenAI pledged to coordinate on detecting and labeling misleading AI deepfakes targeting candidates and voters. Why it matters: With dozens of national elections underway, it's an unprecedented joint effort, though the voluntary accord notably avoids outright bans or binding penalties.
Apple is reportedly expanding internal testing of new generative AI features for Xcode, aiming to release a GitHub Copilot rival to third-party developers later this year. The company is also exploring chatbot-like search upgrades for Spotlight across iOS and macOS to handle complex app actions. After staying relatively quiet in the AI arms race, Apple looks ready to finally show its hand at WWDC.
A federal judge just dismissed nearly all claims in Sarah Silverman and other authors' lawsuits against OpenAI, throwing out allegations of vicarious infringement and DMCA violations. The judge noted plaintiffs failed to show ChatGPT outputs were substantially similar to their works. OpenAI still faces direct infringement claims, but the authors' case just took a massive blow.
Google just announced its next-gen AI model Gemini 1.5, led by 1.5 Pro built on a Mixture-of-Experts architecture. It matches flagship 1.0 Ultra quality with far less compute while shattering context limits with support for up to 1 million tokens—enough for an hour of video or 30,000 lines of code. It's rolling out now in limited preview for developers and enterprise users.
Meta just unveiled V-JEPA, a non-generative video model designed to build human-like world models. Instead of predicting pixels, it predicts masked video parts in an abstract representation space, boosting training efficiency by up to 6x. Best of all, it excels at frozen evaluations—meaning you don't need to retrain the entire model for downstream tasks.
Amazon researchers just unveiled BASE TTS, a 980M-parameter speech model trained on 100,000 hours of public audio that exhibits LLM-like emergent behaviors. Scaling the model allowed it to naturally master tricky tasks like emotional inflection, whispered dialogue, and complex syntax without explicit fine-tuning. Why it matters: Text-to-speech may finally be escaping the uncanny valley, though Amazon is keeping the weights under wraps for safety.
X is testing a new Explore page feature that uses Grok to generate brief summaries explaining why topics are trending. Instead of clicking through vague keywords and hashtags, users get an immediate rundown of the day's key stories. The feature is currently in limited testing and may not roll out globally.
OpenAI just teamed up with Microsoft to terminate accounts tied to five state-affiliated threat groups. The hackers were caught using its models for basic tasks like debugging code, drafting phishing content, and researching open-source vulnerabilities. Why it matters: OpenAI found that AI currently offers only incremental advantages for cyberattacks over existing tools, but shutting them out early keeps the pressure on.
OpenAI founding member and prominent AI researcher Andrej Karpathy has departed the ChatGPT creator. Karpathy, who rejoined OpenAI in 2023 after leading Tesla's Autopilot efforts, noted on social media that his immediate plan is to work on personal projects. His departure marks the loss of one of OpenAI's most influential research voices.
Mozilla just announced it is cutting around 60 jobs and shutting down side projects like 3D platform Hubs to refocus on its flagship browser. The organization is combining its Pocket, content, and AI/ML teams directly into Firefox to build 'trustworthy AI' features. Why it matters: After years of chasing diversified bets, Mozilla is retreating to its core base to survive the generative AI era.
Google just deployed a new machine learning algorithm to catch fraudulent reviews faster than ever. The system analyzes long-term patterns to flag sudden rating spikes and coordinated scam rings, helping block over 170 million policy-violating reviews in 2023—a 45% jump from the previous year.
AI companies just dropped millions on Super Bowl ads to convince a nervous public they are here to help, not steal jobs. With 52% of Americans more concerned than excited about AI, Microsoft and Google used $7 million spots to rebrand their tools as humble companions for everyday tasks like photo-taking and starting a business.
NVIDIA just released Chat with RTX, a free tech demo that lets you run customized chatbots completely offline on Windows. By connecting local files and YouTube links to open-source models like Mistral and Llama 2 via TensorRT-LLM, you can query notes and documents without sending data to the cloud. You will need a GeForce RTX 30 Series GPU or higher with at least 8GB of VRAM.
OpenAI CEO Sam Altman warned that "subtle societal misalignments" — rather than rogue killer robots — are what keep him up at night. Speaking at the World Governments Summit, he reiterated calls for an IAEA-like oversight body while comparing today's AI to early black-and-white cellphones. Why it matters: The industry's top figure insists AI makers shouldn't be the ones writing the rulebook.
The USPTO just issued new guidance declaring that only "natural humans" can be listed as inventors on patents, ruling out AI models as inventors or co-inventors. While AI-assisted inventions aren't barred from patent protection, human applicants must prove they made a "significant contribution" to the conception—such as specific prompt construction or heavily modifying the outputs. Simply running a prompt or owning the model doesn't cut it.
McAfee just revealed that 45% of men plan to use AI to draft Valentine’s Day messages this year, up from 30% last year. While 69% say AI-crafted charm gets better responses, 57% admit they'd be hurt or offended if they found out code wrote their love note.
ElevenLabs just rolled out voice monetization, allowing creators to earn passive cash from their AI voice clones in the Voice Library. Users simply upload 30 minutes of audio to build a clone, getting paid whenever paying subscribers generate speech with it. The platform also includes owner verification and daily payout tracking.
NVIDIA CEO Jensen Huang just told world leaders at Dubai's World Governments Summit that every nation needs to own its own "sovereign AI" to protect local data, culture, and history. Huang urged leaders to build out their own compute infrastructure, arguing that natural language has already turned everyone into a programmer.
Stability AI just launched Stable Cascade in research preview, a text-to-image model built on a three-stage compression architecture that cuts fine-tuning costs by 16x compared to Stable Diffusion. Training, ControlNet, and inference scripts are live on GitHub, though it remains under a non-commercial research license.
Dozens of activists from Pause AI and No AGI just rallied outside OpenAI's San Francisco headquarters, demanding the lab halt work on AGI and urging engineers to quit. The protest followed OpenAI's quiet removal of a ban on military use from its policies and its subsequent work with the Pentagon. Organizers argue tech labs cannot self-regulate and need strict external oversight.
Microsoft just rolled out a new Copilot experience in the latest Windows 11 Canary builds, making the assistant far more context-aware. When you copy text or images, the taskbar icon animates to offer instant actions like summarizing or explaining, while Notepad is also getting right-click AI deciphering for code and logs. It's an early look at how Microsoft plans to reshape Windows for its 2024 "AI PC" push ahead of the 24H2 update.
Disney is beta-testing a new AI tool called "Magic Words" that tags scenes across its catalog by content, brands, and emotional mood. Partnering with six global ad firms, the system lets brands run contextual ads on Disney+ and Hulu tailored to the exact vibe viewers are feeling, rather than relying on broad demographics.
OpenAI CEO Sam Altman is reportedly in talks to raise an eye-popping $5 trillion to $7 trillion to overhaul the global semiconductor industry and expand AI infrastructure. Altman has long warned that chip shortages are throttling AI growth, and he's now pitching sovereign wealth funds like the UAE to help challenge Nvidia's dominance. Why it matters: If realized, it is an unprecedented moonshot that would fundamentally reshape global supply chains.
Perplexity just partnered with Vercel to plug its online search models directly into the Vercel AI SDK. Developers building JavaScript and TypeScript web apps can now tap into pplx-7b-online and pplx-70b-online to fetch live web data without training cutoffs. Vercel also launched a multi-model AI playground alongside the update.
