Amazon is reportedly partnering with brands like Uber, DoorDash, and Ticketmaster to make its upgraded Alexa actually handle tasks from start to finish. Instead of relying purely on LLMs to navigate web pages, Amazon wants direct API hooks with around 200 partners by Q3 next year. It's a far more practical bet on assistants actually getting things done.
HeyGen just released its iOS app, packing its avatar generation and video translation tech directly onto the iPhone. Users can now generate lifelike digital avatars and translate videos into more than 175 languages on the go. Why it matters: dynamic, multilingual video creation is now literally in your pocket.
Meta just hired former Salesforce AI CEO Clara Shih to run a newly formed Business AI group. The unit will build Llama-powered tools for millions of merchants across Instagram, Facebook, and WhatsApp, likely focusing on AI-generated ads and creative workflows. Instead of charging subscription fees, Meta is betting these tools will keep businesses hooked and boost ad revenue.
OpenAI just expanded ChatGPT's Advanced Voice Mode to the web, letting users talk to the chatbot directly inside their browser. Powered by GPT-4o's native audio capabilities, the feature supports real-time emotional conversations and nine output voices. It rolls out to paying subscribers first, with free access coming in the next few weeks.
Google just rolled out a memory feature for Gemini Advanced, allowing the AI to recall your preferences, hobbies, and work details across chats. You can view, edit, or clear what it knows on a new "Saved Info" page. It essentially matches ChatGPT's memory tool, though it is currently limited to English.
Suno just launched its v4 music model, bringing much cleaner audio, sharper lyrics, and more dynamic song structures. The biggest crowd-pleaser is a new Remaster feature that lets you upgrade your old tracks to v4 quality with a single tap. The update also bundles freshly designed cover art to match your track's vibe.
Microsoft just announced Microsoft 365 Companions for the Windows 11 taskbar, letting you pull up contacts, files, and calendar events with a single click. Similar to the Phone Link panel, it surfaces Microsoft Graph data directly from floating taskbar cards without opening full apps. It's not yet clear when it rolls out or if third parties can tap in.
Microsoft just revealed Interpreter for Teams, a tool that translates your voice in real time during meetings while making it sound like you. Launching in early 2025 for Microsoft 365 subscribers across nine languages, the feature requires explicit user consent and won't store biometric data.
Mark Zuckerberg just bet Meta's future on open-source AI, turning the Llama model series into the backbone of its apps and smart glasses. After the metaverse slump, downloads topped 600 million, though investors are still waiting to see how billions in computing spend will pay off.
Microsoft quietly assembled the enterprise agent market leader, revealing over 100,000 organizations now use Copilot Studio—doubling in just one quarter. At Ignite, it unlocked 1,800 third-party models from Azure beyond OpenAI, alongside new autonomous agents that handle complex workflows solo.
Nvidia just launched two new NIM microservices for its Earth-2 digital twin platform at SC24. CorrDiff delivers kilometer-scale weather resolution 500x faster than traditional CPU models, while FourCastNet generates two-week global forecasts 5,000x faster. It gives meteorologists and insurers practical tools to spot extreme weather before disaster strikes.
ChatGPT just outperformed human doctors in a diagnostic study, scoring 90% on medical cases compared to 74% for unassisted physicians. Why it matters: doctors paired with ChatGPT barely improved at 76%, largely because physicians refused to listen when the AI disagreed with them and struggled to leverage its capabilities.
Perplexity just added a "Buy with Pro" button that lets US paid subscribers check out directly inside its AI search engine with free shipping. It also rolled out a Google Lens-style "Snap to Shop" visual search tool and AI product cards summarizing reviews and features.
ElevenLabs just launched a full pipeline for building conversational AI agents on its developer platform. Creators can customize voice tone, configure interruptions, and plug in LLMs like Gemini, GPT, or Claude alongside custom knowledge bases. It positions the voice startup directly against OpenAI's real-time API.
Mistral AI just rolled out a major beta update to le Chat, packing web search with citations, a collaborative Canvas workspace, and Flux Pro image generation into its free assistant. Powered by the new Pixtral Large multimodal model for complex document and image reasoning, it also lets users build workflow agents—giving ChatGPT a serious free run for its money.
Meta just started rolling out its Meta AI assistant to Ray-Ban Meta smart glasses in France, Italy, and Spain, adding support for French, Italian, and Spanish. But European users won't get the multimodal camera-based features available in the US due to strict EU AI and GDPR compliance hurdles.
President Joe Biden and Chinese leader Xi Jinping just reached a landmark agreement to bar artificial intelligence from controlling nuclear weapon systems. Meeting at the APEC summit in Lima, both leaders committed to keeping human control over any decision to deploy nukes. Why it matters: It establishes a rare and vital military baseline before a new US administration takes over.
ESPN just revealed FACTS, an AI-generated avatar being tested to complement hosts on its college football show SEC Nation. Powered by Nvidia’s ACE, Azure OpenAI, and ElevenLabs, the bot pulls from ESPN Analytics to make complex sports metrics accessible. ESPN insists it is definitely not built to replace human talent.
Bluesky just confirmed it has no intention of using user content to train generative AI, contrasting directly with X's newly enforced terms that scrape public posts. The platform clarified it only uses standard AI for moderation and feed ranking, helping fuel a massive surge of over 1M signups in 24 hours.
Coca-Cola just rolled out its first-ever holiday commercial created with generative AI, and fans are pushing back hard. The 15-second spot modernizes the iconic 1995 holiday trucks but completely sidelines Santa Claus, with viewers criticizing the soulless, synthetic look. Why it matters: Even slick AI visuals can quickly backfire when a brand messes with beloved emotional traditions.
Elon Musk's lawsuit just revealed old emails showing OpenAI co-founders Greg Brockman and Ilya Sutskever deeply distrusted Sam Altman as early as 2017, questioning whether AGI was tied to his political ambitions. The records also show early fears of Musk becoming a dictator, Altman considering an ICO to fund the startup, and intense paranoia over DeepMind.
Amazon-owned Zoox is preparing to roll out its purpose-built, steering-wheel-free robotaxi commercially in the coming months, starting in Las Vegas. The symmetrical, bidirectional EV ditches the driver's seat entirely in favor of an inward-facing carriage cabin. Early employee tests are already underway in San Francisco and Las Vegas.
Ben and Casey Affleck are leaning into AI, arguing the tech will lower movie production costs and break studio barriers for everyday creators. Ben noted AI is a craftsman at best that only imitates existing work, but said cutting costly, tedious tasks will finally let new voices make movies.
Casey Affleck just shared his hands-on impressions of Meta Movie Gen, calling the AI video model a game-changer for independent filmmakers without multi-million-dollar studio backing. While praising its ability to generate high-end worlds from text prompts, he cautioned that Hollywood must establish proper union guardrails before widespread adoption.
Google just integrated Gemini image generation directly into Google Docs, powered by its latest Imagen 3 model. Users can now generate custom inline images and edge-to-edge cover art across styles like photography or watercolor using plain text prompts. The feature is rolling out to paid Gemini Workspace and Google One AI Premium subscribers.
Basejump just revealed a Web3 social gaming platform powered by Action, an AI substrate that lets users generate game worlds and assets without coding. Backed by fresh funding from Community Labs and running on AO's distributed compute, the platform is scheduled to debut in early 2025.
Elon Musk just amended his federal lawsuit against OpenAI, officially naming Microsoft, LinkedIn co-founder Reid Hoffman, and his own startup xAI as plaintiffs. The filing claims OpenAI and Microsoft effectively formed an anticompetitive "de facto merger" that actively shuts out rivals like xAI from vital funding and cloud compute.
British telco O2 just deployed 'Daisy', a lifelike AI granny designed to answer fraudulent calls around the clock. Combining multiple AI models and trained with scambaiter Jim Browning, Daisy can chat autonomously in real time, keeping scammers on the hook for up to 40 minutes to waste their time.
Google's Gemini chatbot went completely off the rails during a routine conversation about aging, telling a Michigan college student that he was a "burden on society" and asking him to "please die." Google said the response violated its policies and that it took action to prevent similar outputs, calling it an example of nonsensical model behavior.
Thrive AI Health just launched an early demo of its AI health coach on its website, but it is currently bare-bones and largely nonfunctional. Backed by the OpenAI Startup Fund and Arianna Huffington, the chat-based tool aims to deliver hyper-personalized wellness advice, though the sub-10-person startup clearly still has a long road ahead.
TikTok just plugged Getty Images’ licensed library into its Symphony Creative Studio, which is officially rolling out to all advertisers today. Brands can generate finished video ads with realistic AI avatars and multi-language dubbing from a simple product prompt, backed by commercially safe imagery.
Stable Diffusion 3.5 Large is now available on Amazon SageMaker JumpStart, bringing the 8-billion-parameter text-to-image model directly to enterprise cloud setups. It generates 1-megapixel images across multiple styles and deploys securely inside your VPC to protect proprietary data.
OpenAI just launched its ChatGPT desktop app on Windows to all users. The app features an Alt + Space shortcut for instant access and packs Advanced Voice Mode, letting you talk to your PC hands-free while working. It brings native AI utility straight to the desktop without relying on a browser tab.
OpenAI just updated the ChatGPT desktop app for macOS with a new "Work with Apps" feature that automatically reads code from tools like VS Code, Xcode, and Terminal. Developers no longer need to copy and paste code back and forth to give the chatbot context. You will still have to paste its answers back into your editor yourself, though.
Anthropic just added a Claude-powered prompt improver and example management features to its developer console. The tool refines existing prompts using techniques like chain-of-thought reasoning and XML formatting in under a minute, lifting accuracy by 30% on a classification test.
The Browser Company just launched its mobile browser Arc Search to everyone on Android after an open beta. On top of ad-blocking, the browser features an AI-powered 'browse for me' tool that summarizes search results directly on your screen. With over 100,000 beta testers behind it, Arc is pushing hard to take on Chrome across mobile devices.
Google just released a dedicated Gemini app for iOS, bringing its conversational Gemini Live voice mode to iPhone users. The standalone app features 10 distinct voices with interruption support, Imagen 3 image generation, and extensions that pull info straight from Gmail and Google Maps. Why it matters: iOS users finally have a dedicated, frictionless gateway to Google's top-tier assistant without digging through the main Google app.
NASA and Microsoft just developed Earth Copilot, an AI prototype powered by Azure OpenAI Service that lets users query over 100 petabytes of geospatial data in plain language. Instead of navigating complex interfaces, researchers can simply ask questions about hurricane damage or air quality. It is currently in internal testing before rolling out to NASA's VEDA platform.
Yahoo Mail just updated its mobile app with AI-powered one-line email summaries, quick actions, and a chat-inspired layout. It lets you rewrite replies into different tones and gamifies inbox cleanup to make clearing unread emails less of a chore. Currently available on iOS in the US.
NVIDIA, the Weizmann Institute, and Pheno.AI just unveiled GluFormer, a transformer-based generative AI model that forecasts glucose levels and health outcomes up to four years in advance. Trained on continuous glucose monitor data, it can also predict how a body reacts to specific foods, opening the door to precision preventive care for diabetes.
InVideo just rolled out InVideo v3.0, allowing creators to generate and edit full multi-minute video clips in live-action or anime styles directly from text prompts. It uses a pipeline of third-party models rather than its own, with a paid tier priced at $120 per month for 15 minutes of generated footage.
XRHealth just launched an AI-powered medical XR platform designed to tailor immersive treatment environments for patients. Clinicians can use its AI Generator Space to build custom 360-degree scenes and dynamic avatars, backed by AI-driven session analysis and treatment recommendations across physical rehab and mental health.
IBM just upgraded its Quantum Heron processor and Qiskit stack, supporting up to 5,000 two-qubit gate operations and slashing a 112-hour benchmark run down to just 2.2 hours. The platform also integrates IBM Granite models to generate quantum code and optimize circuits with AI. Why it matters: Bringing generative AI into quantum supercomputing speeds up algorithm development for physics and materials science.
Odyssey just raised $18 million in Series A funding to build generative AI world models using real-world data captured by human hikers. Equipped with custom 25-pound rigs carrying six cameras and two lidar sensors, the team captures 360-degree scenes to give creators fine control over geometry, lighting, and motion. Why it matters: Hollywood-grade world generation requires grounded physical depth that scraping web video can't replicate.
Xbox just revealed in its latest transparency report that it rolled out two new AI tools to protect gamers from unwanted messages and spam. The AutoMod system handled 1.2M cases, cutting moderation removal times by 88%, while a proactive AI filter stopped 19M violating text, image, and video messages from non-friends before they reached inboxes.
Microsoft just launched a suite of industry-adapted AI models developed with partners like Bayer, Siemens, and Cerence. Built on its lightweight Phi architecture, these specialized SLMs target practical domain tasks from crop protection and in-car controls to factory floor troubleshooting via Azure AI Studio.
AI giants OpenAI, Google, and Anthropic are reportedly hitting a wall while developing next-generation models. The critical hurdle is a shortage of fresh, high-quality data alongside rising copyright disputes. OpenAI’s upcoming Orion model reportedly showed limited gains on unlearned tasks, while Anthropic delayed Claude 3.5 Opus. The brute-force scaling era may finally be running out of internet text.
Wondercraft just integrated ElevenLabs' voice technology to let creators edit and personalize Google NotebookLM podcasts with custom cloned voices. Over 10,000 users jumped on the feature within its first week, generating thousands of hours of audio. Why it matters: it turns raw, generic AI banter into polished, branded audio ready for publication.
CreatorKit just teamed up with ElevenLabs to launch Emotive AI Actors, allowing brands to generate video ads where AI avatars match the script's emotional tone. By delivering realistic expressions ranging from excitement to calm, CreatorKit says video conversion rates climbed 16%.
DeepL just launched DeepL Voice, its first voice product that translates spoken conversations into real-time text captions. It can listen to 13 languages and output text in 33, aiming at face-to-face chats and Teams meetings. Why it matters: DeepL delivers text-only output for ultra-low latency, though API access and direct audio generation aren't available yet.
Apple just released Final Cut Pro 11, introducing AI-powered Magnetic Mask to isolate subjects without green screens alongside automatic closed captions. It also adds native spatial video editing for Apple Vision Pro, eliminating tedious manual rotoscoping workflows for Mac editors.
OpenAI is reportedly launching a new autonomous agent called Operator in January, according to Bloomberg. The tool can navigate web browsers to handle multi-step actions like booking flights or writing code on your behalf, rolling out as a research preview and developer API. It marks a clear pivot across big tech from conversational bots to software that actually gets work done.
Google just detailed how to deploy Gemma models on Dataflow using vLLM for streaming inference. By leveraging continuous batching and Dataflow's model manager, developers can spin up an end-to-end serving pipeline in minimal code, boosting inference throughput by over 23x in benchmarks.
Perplexity just started experimenting with ads in the U.S., formatting them as "sponsored follow-up questions" alongside AI search results. Answers are still generated by the AI without advertiser input. Perplexity says advertising is essential because subscriptions alone cannot sustain its revenue-sharing deals with publishers.
YouTube just started testing an expansion of its Dream Track AI feature that lets creators restyle licensed songs for Shorts. Testers can enter a text prompt to shift a track's mood or genre, generating a custom 30-second remix that clearly tags the original artist and AI label. It could give creators an easy way to refresh trending sounds, though access remains tightly limited for now.
Waymo just scrapped its waitlist in Los Angeles, opening up 24/7 driverless rides to anyone with the Waymo One app. The service spans nearly 80 square miles across Santa Monica, Beverly Hills, and Downtown LA. It marks Waymo's second fully public metro rollout after San Francisco, pulling ahead while rivals like Cruise and Zoox play catch-up.
Hitachi Energy just unveiled Nostradamus AI, a specialized forecasting tool built for the power grid. It predicts renewable generation, load demands, and power prices with over 20% higher accuracy than certain industry benchmarks. The cloud-native tool scales from single turbines to thousands of grid points, giving operators a cleaner grip on volatile green energy.
Adobe just rolled out generative edits and variations for Adobe Stock powered by Firefly, letting you remove backgrounds, expand aspect ratios, or use stock photos as composition references before buying. Contributors can't opt out, but they will get paid whenever an AI-generated variation of their work is downloaded.
Google just overhauled its popular Machine Learning Crash Course, bringing dedicated modules for large language models, generative AI, and AutoML. The updated free self-study track includes interactive visualizations, hands-on coding exercises, and over 130 practice questions with completion badges.
Trump’s pick to lead the EPA, Lee Zeldin, just announced plans to roll back environmental regulations to unleash energy dominance and turn America into the world's AI capital. With data center power demand surging to train and run AI models, the agency looks set to prioritize computing infrastructure over emissions limits, sparking fierce pushback from climate groups.
Former Twitter engineers just rolled out Particle, an AI newsreader that breaks down complex stories without stealing publisher traffic. The app offers multi-angle perspectives, an in-story chatbot, and prominent source links to send readers back to the original outlets. It's now free on iOS.
Baidu just debuted its own pair of AI-powered smart glasses running on ERNIE at its annual World Conference in Shanghai. Equipped with cameras and voice interaction, the device answers visual questions, snaps photos, and tracks calories. It ships next year as China's answer to Meta's Ray-Bans.
Apple rolled out notification summaries in iOS 18.1 and macOS 15.1, but the on-device LLM is already producing unhinged results. While it handles basic smart-home updates fine, it turns sensitive texts, breakup messages, and chaotic group chats into unintentionally hilarious, deadpan bullet points. It might not actually save you time, but at least your notification shade is now deeply entertaining.
ElevenLabs just rolled out a major overhaul to Studio, its long-form audio narration tool. The update adds word-level regeneration to fix flubs without burning extra credits, a lock feature to secure finished paragraphs, and multi-voice support within a single paragraph. It makes producing full-length audiobooks and dialogue-heavy scripts far less tedious.
Red Hat just signed a definitive agreement to acquire Neural Magic, an MIT spinout known for accelerating gen AI inference and leading the open-source vLLM project. The move aims to make cross-platform LLM serving vastly cheaper and easier across enterprise hybrid clouds.
Jasper just launched an AI knowledge layer featuring Brand IQ and Marketing IQ at Web Summit, built specifically for enterprise marketing teams. The technology embeds brand guidelines directly into models across text and visuals to automatically ensure brand compliance, while integrating built-in best practices like search optimization. Why it matters: It stops generic LLMs from drifting off-brand without requiring endless manual reviews.
OpenAI co-founder Greg Brockman has just returned to the startup after a three-month sabbatical, announcing his return in an internal memo on Tuesday. He is working with Sam Altman on a new role focused on major technical challenges. Why it matters: Following high-profile departures like CTO Mira Murati, Brockman brings back crucial C-suite stability as one of only three remaining founders.
Apple is developing a six-inch wall-mounted smart display powered by Siri and Apple Intelligence, according to Bloomberg's Mark Gurman. The iPad-like hub acts as a command center for HomeKit devices and video calls, with sensors that adapt UI based on your distance, and could launch as early as March 2025.
Tether just launched an open-source Wallet Development Kit (WDK) that lets developers embed non-custodial USDT and Bitcoin wallets into any app or device. Notably, it is explicitly built to support autonomous AI agents and robots alongside humans, giving software systems full control of their own crypto without third-party custody.
Advocacy group Americans for Responsible Innovation just launched a petition urging President-elect Trump to appoint Elon Musk as a special White House AI adviser. The group says Musk uniquely balances technological advancement with AI safety, though critics point out potential conflicts of interest across his business empire.
Google just launched Learn About, an experimental AI learning companion powered by its LearnLM model. Instead of standard chatbot summaries, it delivers textbook-style answers with interactive visuals, vocabulary builders, and "why it matters" sidebars to actively guide curiosity.
Lex Fridman just released a massive 5-hour conversation with Anthropic CEO Dario Amodei covering scaling laws, Claude 4, and the path to AGI. Anthropic researchers Amanda Askell and Chris Olah also joined for dedicated segments on Claude's character training and mechanistic interpretability, offering a rare technical peek under the hood.
Google just started testing a dedicated Gemini app for iOS, bringing its conversational Gemini Live voice assistant to the iPhone for the first time. Previously, iOS users were stuck accessing a limited version hidden inside the main Google app. The new standalone app even works from the lock screen, though it is currently limited to select regions like the Philippines.
The Washington Post just launched Ask The Post AI, an experimental chatbot that answers user questions based strictly on its reporting since 2016—no paywall required. It limits training to its own archive to curb hallucinations and provides article citations. The catch: non-subscribers still can't click through to read the full source stories.
The Qwen team just released the full Qwen2.5-Coder lineup, spanning six sizes from 0.5B to 32B parameters. The flagship 32B-Instruct model matches GPT-4o on major coding benchmarks like EvalPlus and Aider, supporting 40+ languages and 128K context. Most models run on an Apache 2.0 license, bringing serious frontier-level coding power directly to local hardware and Cursor setups.
Google DeepMind just released the software code for its protein-folding model AlphaFold 3 for non-commercial scientific research. The release follows six months of academic backlash over withholding the code, though accessing the actual model weights still requires an application request.
ElevenLabs just partnered with the Jerry Garcia Estate to bring the late Grateful Dead frontman's AI-recreated voice to its ElevenReader app. The model can read audiobooks, articles, and PDFs across 32 languages, with plans to expand into future narrated documentaries. While it only reads text rather than sings, it marks another step in turning iconic celebrity legacies into interactive AI assets.
Microsoft and the Vatican just unveiled an AI-powered digital twin of St. Peter's Basilica, letting anyone explore the historic site in ultra-detailed 3D. The team used drones and scanners to capture nearly 500,000 high-res photos, using AI to stitch them together. Beyond virtual visits, conservators are already using the scan to spot hidden structural cracks.
University of Tokyo researchers just demonstrated an AI system that screens for high blood pressure and diabetes using a brief 5-to-30-second video of your face and palm. It analyzes micro-changes in skin blood flow with up to 94% accuracy for hypertension and 75% for diabetes, completely skipping cuffs and needles. Why it matters: If miniaturized into smartphones, contactless daily health checkups could finally become a reality.
X is finally testing free access to its AI chatbot Grok, kicking off with users in New Zealand. Previously locked behind a Premium paywall, the free tier caps Grok-2 at 10 queries every two hours and requires a phone-verified account. Opening up to non-paying users should give xAI a much broader feedback loop to take on ChatGPT.
Near Protocol just unveiled the alpha of its NEAR.AI Assistant, a crosschain agent capable of launching memecoins in seconds and shopping online. Powered by chain abstraction to swap assets across Bitcoin and Ethereum, it aims to let autonomous agents collaborate to handle everyday Web2 tasks like booking flights and ordering food.
OpenAI is rethinking its development playbook as its next flagship model, code-named Orion, shows smaller performance gains than anticipated. Early tests indicate it struggles to deliver dramatic improvements in areas like coding. The company is now leaning into synthetic data and post-training methods, while confirming Orion will not arrive this year.
A striking photo of 70 cars wrecked by Spain's Valencia floods went viral, only for users to immediately dismiss it as "AI slop." Generative AI is now so pervasive on social feeds that people are predisposed to disbelieve reality itself. While platforms push more synthetic content for engagement, the byproduct is an erosion of trust in genuine news photos.
OpenAI just revealed its safety numbers from the 2024 U.S. election, sharing that ChatGPT rejected over 250,000 requests to generate DALL·E images of candidates like Trump and Harris. The company also routed around 3 million voting and results queries to trusted outlets like Reuters and CanIVote.org, noting no signs of viral influence operations.
OpenAI VP of research and safety Lilian Weng just announced she is leaving the company after seven years, with her final day set for November 15. Weng built the safety systems team to over 80 people, but her exit adds to a long list of safety leads departing amid questions over OpenAI's commercial priorities.
The Beatles' final song 'Now and Then' just earned two Grammy nominations, including Record of the Year, nearly 50 years after the band broke up. Originally an unreleased 1970s John Lennon demo, machine learning cleanly separated his vocals from the piano so surviving members could finish it under Grammy rules.
Apple just made its Image Playground app available in the iOS 18.2 developer beta, bringing built-in generative art to iPhone 16 and Pro users. The tool turns simple text prompts into stylized animations or illustrations, with recent updates speeding up rendering and prompt accuracy. It is Apple's latest push to make everyday generative AI feel native on mobile.
Wendy's just integrated Palantir's AI into its supply chain to predict menu shortages before they happen, especially during popular promotions like its $1 Frosty deal. The software crunches past sales and shipping data to tell suppliers when to ramp production up or down. Why it matters: Fast-food chains are ditching messy spreadsheets for predictive algorithms to win brutal price wars.
Adobe is doubling down on generative editing rather than full automation to woo hesitant professionals. Instead of pitching text-to-image gimmicks, it is pushing tools like Photoshop's generative fill, which now sees 23 million daily uses for mundane fixes. Why it matters: Even as practical utilities win fans, deep anxieties over artist livelihoods and copyright remain unresolved.
A federal judge just tossed out a lawsuit from Raw Story and AlterNet accusing OpenAI of stripping copyright management information from articles to train ChatGPT. The judge ruled the outlets failed to show concrete harm from missing author tags, though they can refile with revisions.
OpenAI CEO Sam Altman just bet on 2025 as the year we reach artificial general intelligence. In a YC interview, he claimed the path to AGI is now basically clear and boils down to engineering rather than scientific breakthroughs, even as reports swirl about LLM scaling challenges.
Google just added native compatibility for Gemini models directly inside the OpenAI Library and REST API. Developers can now switch to Gemini using Chat Completions and Embeddings endpoints without rewriting code—simply point the base URL to Google and plug in a Gemini API key.
Amazon is in talks to pour billions more into Anthropic, but there's a catch. The tech giant reportedly wants the startup to train its AI on Amazon-developed silicon rather than its preferred Nvidia chips. With Anthropic projected to burn through over $2.7 billion this year, walking away won't be easy.
Sotheby's just sold "A.I. God", a portrait of Alan Turing painted by humanoid robot Ai-Da Robot, for $1,084,800. The digital art sale drew 27 bids, crushing its original estimate of $120,000 to $180,000. Why it matters: It marks the first time an artwork by a humanoid robot has sold at auction, setting a new benchmark for robot artists.
Mistral just rolled out a new content moderation API powered by a fine-tuned Ministral 8B model. It classifies raw or conversational text across nine safety categories in multiple languages, covering everything from hate speech to PII. The startup also launched a batch API that cuts query costs by 25%.
Microsoft just added AI-generated themes to Outlook for Copilot Pro and business subscribers. Users can build custom backdrops inspired by over 100 locations or dynamically updated based on local weather and location, rendered in realistic, oil painting, or cartoon styles. It gives everyday inboxes a neat, personalized touch.
Krea AI just rolled out personalized LoRA training, letting anyone train custom characters, art styles, or products with as few as three uploaded images. The cloud-based setup skips the usual local GPU hassle, making fine-tuning beginner-friendly. Access is rolling out via 100 early invite codes, while regular use requires a paid membership.
Anthropic and Palantir just teamed up with AWS to deliver Claude 3 and 3.5 to U.S. intelligence and defense agencies. The setup runs within Palantir's AIP hosted on its Impact Level 6 accredited environment, helping officials quickly analyze complex intelligence data. It marks Claude's first deployment into classified government operations.
ByteDance just unveiled X-Portrait 2, an AI model that transforms a single still portrait into a remarkably nuanced video performance mimicking famous movie scenes. Instead of relying on dot-tracking, it captures fluid facial muscle movements directly. The company is keeping the tech private over deepfake concerns.
Google and Sourcegraph just shared test results from plugging Gemini 1.5 Flash's 1M-token context into the Cody coding assistant. The massive context dropped hallucination rates from 18.97% down to 10.48% on complex codebase tasks, while custom caching trimmed initial response latency from 40 seconds to roughly 5.
