OpenAI just confirmed that its upcoming DevDay won't feature its next major flagship model. Instead of a splashy keynote, the company is shifting to a three-city roadshow across San Francisco, London, and Singapore to focus primarily on API updates and developer sessions.
A YouTube creator just filed a class-action lawsuit against OpenAI, alleging the company secretly transcribed millions of YouTube videos to train ChatGPT and GPT-4 without consent, seeking over $5 million in damages. With high-quality web text rapidly drying up, video transcripts have become a crucial data well—and now, the latest front in AI copyright battles.
Nvidia has been scraping massive amounts of video from YouTube, Netflix, and other platforms to train its AI models, according to leaked internal documents obtained by 404 Media. Employees raised copyright concerns, but managers insisted the project—code-named Cosmos—had approval from top leadership. Nvidia defended the practice as fully compliant with copyright law, while quietly developing 3D world generators and digital humans behind closed doors.
Elon Musk just revived his legal battle against OpenAI, filing a new federal lawsuit against Sam Altman and Greg Brockman. The suit claims Musk was conned into funding a fake non-profit and takes direct aim at OpenAI's partnership with Microsoft. Musk is basically arguing that Altman's entire pivot to profit was an elaborate grift from day one.
Nvidia just unveiled a suite of simulation tools and NIM microservices that let developers teleoperate humanoid robots using Apple's Vision Pro. By tracking hand gestures through the headset, Nvidia's MimicGen service generates synthetic training data from minimal demonstrations, slashing the time and cost needed to train models like Project Groot.
AI chip startup Groq just pulled in a massive $640 million funding round led by Blackrock, valuing the inference specialist at $2.8 billion. The company claims its custom LPUs run generative AI models 10x faster and with a fraction of the energy of conventional GPUs. Meta's Yann LeCun is also joining as a technical advisor, giving the upstart serious ammo against Nvidia.
OpenAI built a text watermarking tool capable of catching students who use ChatGPT for assignments, but is now hesitating to release it. The company warned the invisible watermark can be easily bypassed using translation tools or other AI models, and could unfairly stigmatize non-native English speakers. Why it matters: Detecting AI text remains an unsolved mess, and OpenAI is worried its fix might do more harm than good.
Nvidia has reportedly delayed its next-gen Blackwell B200 chips by at least three months due to a design flaw discovered late in production. Massive shipments to heavyweights like Microsoft, Meta, and Google are now pushed into early next year. Nvidia says production will still ramp in the second half, but declined to comment on rumors.
Meta is reportedly negotiating multi-million-dollar deals with actors like Judi Dench, Awkwafina, and Keegan-Michael Key to license their voices for its Meta AI assistant. The company aims to wrap up deals ahead of its September Connect event, bringing the voices to Instagram and Ray-Ban smart glasses. Why it matters: Meta is betting that authentic star voices will make hardware like smart glasses feel less like a gadget and more like an engaging companion.
Google just pulled its "Dear Sydney" Olympic commercial following days of intense backlash. The ad showed a father using Gemini to write a fan letter for his young daughter to her track idol, sparking outrage over AI replacing authentic human emotion. Google claimed it was only meant to illustrate early drafting, but viewers felt it completely missed the point.
Google just brought back Character.AI co-founders Noam Shazeer and Daniel De Freitas to join DeepMind, alongside signing a non-exclusive licensing deal for the startup’s tech. Most staff will remain at Character.AI under an interim CEO, with the startup pivoting toward third-party models. It is another big-tech reverse acqui-hire that regulators are already eyeing closely.
Adobe Illustrator now lets users generate editable vector graphics, fill complex shapes, and build seamless patterns directly from text prompts. Because the output is true vector art rather than pixels, every element can be ungrouped and scaled infinitely. Why it matters: It bridges the gap between rough ideas and production-ready design files, even if the node logic still needs occasional manual cleanup.
OpenAI-backed Figure just teased Figure 02, its next-generation humanoid robot set to debut on August 6, 2024. The teaser shows off sleek new joints, flexible mesh body designs, and torques reaching 150Nm. Founder Brett Adcock boldly claims it will be the most advanced humanoid robot on the planet.
Google just rolled out three AI-powered features for Chrome built on its Gemini models. Desktop users can now tap Google Lens directly from the address bar to select and query web content, while an upcoming Tab compare tool compiles product comparison tables across tabs, alongside natural-language history search.
AI music platform Suno just fired back at the major record labels' copyright lawsuit, admitting it trains on open-web music while arguing that learning styles and patterns isn't infringement. Suno compared its training to a kid learning rock by listening to records and said it blocks direct artist imitation. The case is set to become a massive legal test for fair use in generative AI.
Stability AI just launched Stable Fast 3D, a rapid reconstruction model that creates complete 3D assets from a single image in just 0.5 seconds. Built on TripoSR, it slashes inference time from SV3D's 10 minutes down to sub-second speeds while generating meshes and material parameters. Weights and code are open-sourced on Hugging Face and GitHub under a community license allowing commercial use for sub-$1M revenue teams.
Black Forest Labs just launched with a $31M seed round led by Andreessen Horowitz and released FLUX.1, a new 12B-parameter text-to-image suite. Created by researchers behind Stable Diffusion, the suite features pro, dev, and schnell versions, delivering top-tier prompt adherence and visual detail that surpasses Midjourney v6.0. Next on the roadmap: text-to-video models.
OpenAI just started rolling out its advanced voice mode to a select group of ChatGPT Plus subscribers, and early demos are popping up online. Videos show the bot imitating pilot intercoms, correcting French pronunciation, and mimicking regional US accents. OpenAI confirmed the feature will expand to all paying Plus users this fall.
Anthropic just made Claude officially available across Brazil, giving users and developers direct access via web, mobile apps, and the API. Paid Pro plans run R$ 110 per month with full Claude 3.5 Sonnet access. Why it matters: Anthropic is speeding up global expansion to meet surging regional demand for frontier models.
Helen Toner, the former OpenAI board member who helped engineer Sam Altman's brief firing, is now building a profile as Washington's trusted AI policy expert. After surviving Silicon Valley backlash and dodging Elon Musk's subpoena drama, she is focused on shaping AI safety and governance in the capital.
OpenAI just agreed to give the U.S. AI Safety Institute early access to its next frontier model for pre-deployment testing. The move aims to push back against mounting criticism over sidelined safety researchers and gutted internal guardrails. Details on how much access regulators will actually get remain scarce.
Google just faced intense backlash over its new Olympics ad for Gemini, titled "Dear Sydney." In the commercial, a father asks the chatbot to write a fan letter to a track star on behalf of his young daughter, prompting viewers to criticize it for outsourcing heartfelt human expression. The outrage highlights a growing public discomfort when tech companies suggest AI should replace authentic creativity and emotion.
Researchers at USC just used machine learning to solve a deadly medical puzzle: reprogramming glioblastoma cells into immune-activating dendritic cells. By turning the cancer cells into antigen-presenting guides inside the tumor, the method bypasses the blood-brain barrier and boosted mouse survival rates by up to 75% alongside immunotherapy. Why it matters: It offers a biological Trojan horse against one of humanity's deadliest cancers.
Microsoft just officially added OpenAI to its competitor list in its latest annual report, specifically flagging AI offerings alongside search and news advertising. Even though Microsoft poured $13B into the startup, OpenAI's new SearchGPT shows they're now stepping straight onto each other's turf. OpenAI says the partnership remains intact.
Tech-loaded "AI toothbrushes" from giants like P&G, Colgate, and Philips might not actually improve your oral health, according to an Ars Technica investigation. Instead, the embedded sensors and companion apps serve as a direct pipeline for brands to capture real-world brushing habits, feed internal R&D, and pitch personalized product ads. Why it matters: The AI hype wave has officially entered your bathroom, and your daily hygiene routine is now prime behavioral data.
Leonardo.Ai just detailed how its upcoming in-house foundation model, Phoenix, aims to give creators precise prompt adherence and granular control. Fresh off being acquired by Canva, the 17-million-user platform is moving beyond open-source Stable Diffusion roots to replace prompt roulette with deliberate visual creation.
Yum! Brands just announced plans to expand its "Voice AI" ordering system to hundreds of Taco Bell drive-thrus in the US by the end of 2024. The tech has been tested across 100+ stores for two years to cut wait times and ease worker workload. Why it matters: While McDonald's recently retreated from drive-thru AI tests, Yum! is aggressively pushing forward.
Google just rolled out its "About this image" tool to Circle to Search and Google Lens on Android and iOS. Users can now circle any photo on their screen or scan it in the Google app to inspect its web history, metadata, and DeepMind SynthID watermarks for AI generation. Why it matters: fact-checking questionable images is now a one-tap habit across 40 languages.
Vimeo just launched an AI-powered translation tool that translates videos, audio, and subtitles into 29 languages while cloning the speaker's original voice. It accepts video inputs in over 50 languages, and viewers can toggle between translated audio tracks right inside the player. Why it matters: It bypasses expensive voice actors and agencies, making global corporate training and marketing videos far cheaper to localize.
Runway just announced Gen-3 Alpha Turbo, a new video model that generates clips seven times faster than the original Gen-3 Alpha. CEO Cristóbal Valenzuela noted it can render a 10-second clip in just 11 seconds at a lower cost, rolling out to users over the coming days.
Google just released Gemma Scope, a massive open-source suite of over 400 sparse autoencoders to help researchers decipher Gemma 2's inner workings. By isolating more than 30 million learned features across layers, the tool functions like a microscope for mechanistic interpretability. Why it matters: Cracking open neural network black boxes is key to catching hallucinations, deception, and safety vulnerabilities before they cause harm.
OpenAI just threw its weight behind three U.S. Senate bills, notably supporting the Future of AI Innovation Act to formally authorize the U.S. AI Safety Institute. The company is actively courting lawmakers ahead of upcoming regulatory scrutiny to secure a front-row seat at the policy table.
Meta just shut down its celebrity lookalike AI chatbots after spending millions on likeness rights just last September. Instead of handcrafted personas based on stars like Dwyane Wade and Paris Hilton, the company is pivoting to AI Studio, letting creators build their own bot personas.
Activision just released an open-source dataset pulled from Call of Duty: Warzone’s Caldera map on GitHub. Formatted in OpenUSD, it includes near-complete map geometry and anonymized player movement traces to support research. It marks one of the largest production-validated gaming datasets ever made public, giving generative AI models rich 3D environments to learn from.
Midjourney just launched V6.1 as its new default image model. The update delivers cleaner hands, faces, and textures, smarter text rendering, and a 25% boost in standard generation speeds, alongside an upgraded 2x upscaler. Inpainting and zooming still temporarily fall back to V6.0.
Avi Schiffmann just unveiled Friend, a $99 screenless AI pendant built solely for emotional companionship. The Bluetooth necklace constantly listens to your day and texts you like a real friend rather than helping with work, backed by $2.5M in funding with shipments slated for January 2025.
The White House just pushed back against clamping down on open-source AI. A new Commerce Department report found no need to restrict developers from releasing model weights, citing insufficient evidence of major harms and real benefits to open innovation. But officials warn they're still monitoring safety and will step in if bigger risks appear.
Perplexity just launched a revenue-sharing program for publishers, signing up partners like TIME, Der Spiegel, and Fortune. When media content is cited in sponsored follow-up queries, publishers will earn a revenue share, plus free API access to build custom on-site answer engines and a year of Enterprise Pro for their entire staff.
Canva just acquired generative AI startup Leonardo.Ai to bake advanced image generation directly into its design suite. Leonardo.Ai will continue to operate as an independent platform with unchanged pricing, while its foundational models power new creative features across Canva.
OpenAI just launched an experimental model dubbed gpt-4o-64k-output-alpha that can generate up to 64K tokens in a single completion. It targets new long-form workflows, though higher inference costs push pricing to $6 per million input tokens and $18 per million output tokens for Alpha testers.
The US Air Force is funding five mystery companies to develop autonomous AI brains for its Collaborative Combat Aircraft (CCA) fleet, keeping the roster classified to block prying eyes. The software will eventually fly drones built by General Atomics and Anduril, with a final production bake-off scheduled for 2026.
Meta just rolled out AI Studio in the US, allowing anyone to build and discover custom AI characters powered by Llama 3.1. Creators can set up AI extensions of themselves to answer common questions and engage with followers. However, Meta later noted it has halted new character creation as it shifts focus to Muse Spark.
NVIDIA just showcased James, an interactive digital human built on its ACE platform and ElevenLabs voice tech that connects with users using humor and emotion. The company also previewed Maxine 3D, which turns 2D video feeds into real-time 3D avatars for telepresence. Virtual brand ambassadors are quickly shifting from uncanny bots into expressive, lifelike agents.
Runway just rolled out Image-to-Video capabilities for its Gen-3 Alpha model, allowing creators to anchor any static image as the first frame alongside text prompts. The update delivers much tighter artistic control and visual consistency, giving creators a reliable starting point instead of rolling the dice with pure text.
Elon Musk just shared an AI-manipulated parody of a Kamala Harris campaign ad on X without context, calling it "amazing." By omitting the original creator's parody disclaimer, the repost appears to breach X's own rules against deceptive synthetic media, drawing fierce pushback from lawmakers.
Shutterstock and Getty Images just rolled out major visual AI upgrades powered by Nvidia’s Edify architecture. Shutterstock launched a commercial beta for Generative 3D that creates editable assets and 16K HDRi lighting maps from prompts. Meanwhile, Getty doubled its generation speed to four images in six seconds and added fine-tuning, both leveraging commercially licensed data.
NVIDIA just announced a full suite of services, models, and computing platforms for humanoid robot developers at SIGGRAPH. Key additions include new NIM microservices for synthetic motion data, an OSMO orchestration cloud service that cuts development cycles from months to under a week, and a teleoperation workflow using Apple Vision Pro to turn tiny amounts of human demonstration into massive training data. Early access is now open via NVIDIA's developer program.
Apple is reportedly delaying Apple Intelligence past the initial September release of iOS 18, pushing public rollout to an October software update. That means the first batch of iPhone 16 buyers won't get AI out of the box, as Apple buys more time to squash bugs. Developers are expected to get early access via an iOS 18.1 beta this week.
AMD just teamed up with TensorStack to release Amuse 2.0 Beta, a local AI suite that runs Stable Diffusion and ControlNet workflows without any command-line setup. The biggest perk is XDNA Super Resolution, which speeds up creation and doubles output resolution on supported Ryzen AI processors and Radeon RX 7000 GPUs. It gives PC users a seamless on-device sandbox that bypasses the cloud entirely.
The IOC's Olympic AI Agenda is seeing its first real-world test at the 2024 Paris Olympics. AI tools are assisting coaches with real-time motion analysis and injury prevention, supporting gymnastics judging, and automatically scrubbing abusive social media posts directed at athletes. Why it matters: Elite sports are treating AI not just as a training shortcut, but as infrastructure for competitive fairness and athlete safety.
Elon Musk’s X quietly turned on a default setting allowing user posts and interactions to train xAI’s Grok chatbot, sparking immediate pushback from UK and Irish regulators. The pre-ticked consent setting breaches GDPR rules against default opt-ins. Users currently have to dig into the web version's settings just to turn it off.
Apple just signed the White House's voluntary commitment to AI safety, joining 15 other tech giants including OpenAI and Google. Under the pledge, Apple agrees to conduct pre-release red-team testing, treat model weights confidentially, and develop content watermarking systems. With Apple Intelligence rolling out to 2 billion devices soon, the move is a clear effort to appease regulators early.
Google just upgraded the free version of Gemini to its speedier 1.5 Flash model, bringing noticeable latency drops, stronger reasoning, and a 4x context boost to 32K tokens. Responses now include clickable source chips to tackle hallucinations, with file uploads and chart visualization rolling out soon. Why it matters: Google is making its fastest frontier capabilities accessible to everyone without a paywall.
Anthropic just came under fire after its ClaudeBot crawler pounded repair site iFixit nearly a million times in 24 hours, ignoring terms of service that explicitly ban AI scraping. Anthropic defended itself by claiming its bot only listens to robots.txt files, which iFixit eventually used to stop the assault. Why it matters: It shows how toothless website policies are when AI giants insist on robots.txt as the only off switch.
Runway just got caught in a major training data leak after 404 Media obtained an internal spreadsheet showing the company scraped thousands of YouTube channels and pirated films to train its flagship Gen-3 video model. Why it matters: Despite raising $141M from Google and Nvidia while dodging questions about data provenance, Runway now faces intense scrutiny as AI copyright tensions boil over.
AlphaProof and AlphaGeometry 2 just solved four out of six problems from the 2024 International Mathematical Olympiad, achieving a silver-medal standard for the first time. The hybrid systems scored 28 out of 42 points, including cracking the contest's hardest problem. It marks a major milestone toward building AI systems capable of complex mathematical reasoning.
OpenAI just unveiled SearchGPT, a temporary prototype designed to give fast, timely answers backed by clear web sources. Users can ask conversational follow-up questions with shared context, and sites can appear in results even if they opt out of AI training. It’s OpenAI’s clearest signal yet that search is shifting from ten blue links to direct dialogue.
Kuaishou just rolled out Kling AI 1.0 worldwide, opening its Sora competitor to anyone with an email address. Users receive 66 free daily credits for text-to-video and image-to-video generation, complete with multi-format outputs and custom camera controls. While OpenAI keeps Sora behind closed doors, Kling is giving creators real access right now.
Microsoft just started testing a radical Bing redesign that puts AI-generated answers front and center while banishing traditional blue links to a skinny right sidebar. The new layout turns search into multi-section deep dives with charts and videos. Microsoft claims it maintains publisher clicks, but it's currently limited to a handful of queries.
Reddit is now blocking web crawlers and search engines from indexing its recent posts unless they pay up. Google is currently the only major engine surfacing fresh Reddit results, following its $60 million AI training deal, while Bing and DuckDuckGo are left in the cold after Reddit updated its robots.txt file.
Stability AI just launched Stable Video 4D, an open-source model that turns a single video into eight novel-view videos. It generates 5 frames across 8 camera perspectives in about 40 seconds, streamlining dynamic 3D optimization for game development and video editing. The weights are live on Hugging Face.
X just rolled out a Grok-powered "More About This Account" button for paid web users, letting you hover over any profile to generate a quick AI summary. The problem? Early tests show Grok frequently confuses users with celebrities or invents hobbies out of thin air. Why it matters: Even with direct access to platform data, turning social posts into reliable user context is proving harder than expected.
Mistral AI just dropped Mistral Large 2, a 123B-parameter flagship model built with a 128k context window designed to run efficiently on a single node. It delivers major gains in coding and math that rival GPT-4o and Claude 3 Opus, while actively curbing hallucinations. Non-commercial weights are free to download, offering serious frontier power without giant infrastructure.
OpenAI just shared details on Rule-Based Rewards (RBRs), a safety alignment method that uses clear-cut logic instead of massive human feedback. By scoring outputs against step-by-step guidelines, it weeds out harmful content and reduces false refusals on benign prompts. Why it matters: Safety policies can now update on the fly without costly retraining loops.
Suno just launched a new "Stems" feature for Pro and Premier subscribers, allowing users to split AI-generated tracks into separate vocal and instrumental files. Each stem generates as an individual clip that can be downloaded or extended independently. It gives creators far more control over mixing, editing, and repurposing their AI tracks.
Senate Democrats just sent a pointed letter to OpenAI CEO Sam Altman, demanding the company commit to letting U.S. agencies test its next frontier model before public release. Lawmakers also pressed OpenAI on compute allocation for safety following whistleblower complaints over rushed launches. Regulators are officially running out of patience with voluntary self-policing.
Cohere just launched Rerank 3 Nimble, a new foundation model built for enterprise search and RAG workflows. It delivers roughly 3x faster performance than Rerank 3 while keeping high retrieval accuracy, but it is currently exclusive to Amazon SageMaker.
NVIDIA just broke down how its AI Foundry service helps companies build custom generative AI models using proprietary data. Powered by DGX Cloud, the NeMo platform, and support for Llama 3.1, it lets businesses forge and deploy tailored models as NIM microservices without building infrastructure from scratch.
Meta just rolled out a major upgrade to Meta AI, adding preview access to its flagship Llama 3.1 405B model alongside a new "Imagine me" stylized selfie feature. Users can now generate custom scenes starring themselves, though the 405B model requires manual switching and has weekly limits.
Meta just dropped its open-source Llama 3.1 family, headlined by a massive 405B model that rivals GPT-4o and Claude 3.5 Sonnet across key benchmarks. The entire lineup supports a 128K context window, and Meta has updated its license so developers can finally use 405B outputs to distill smaller models.
Adobe just rolled out major AI updates across Illustrator and Photoshop. Illustrator gained Generative Shape Fill, Text to Pattern, and Mockup tools to quickly generate vector graphics and 3D product previews. Meanwhile, Photoshop officially launched its Generate Image tool to general availability alongside new precision selection brushes. Why it matters: Adobe is embedding generative AI deeper into everyday creative workflows, letting pros quickly iterate concepts without leaving their canvas.
