News/Meta
Meta’s Ray-Ban glasses to gain 4 AI-powered features — and memory
Meta unveils groundbreaking features for Ray-Ban smart glasses: Meta's latest update to their Ray-Ban smart glasses introduces innovative AI-powered capabilities, enhancing user experience and accessibility. Multimodal AI integration: The smart glasses now leverage advanced multimodal AI technology, allowing for more natural and intuitive interactions with the device. Users can now engage with the AI assistant without using the "Hey Meta" wake phrase, enabling seamless conversation and queries about their surroundings. The glasses can analyze visual and auditory input in real-time, providing context-aware responses to user inquiries. This enhancement allows users to ask questions about objects or situations they encounter without...
read Sep 25, 2024AI privacy hoax spreads on Facebook: what users need to know
The rise of AI-related privacy concerns: A viral copypasta claiming to protect users from Meta's AI has spread across social media platforms, highlighting growing anxieties about personal data and AI training. The "Goodbye Meta AI" copypasta, featuring legal-sounding language and urgent calls to action, has been shared thousands of times on Facebook and Instagram. Notable figures, including Tom Brady and James McAvoy, have reportedly shared the copypasta in their Instagram Stories. This trend echoes similar viral posts from over a decade ago, which also claimed to protect users' data through pseudo-legal declarations. Debunking the copypasta's claims: Despite its widespread circulation,...
read Sep 24, 2024Meta’s AI Shades Could Revolutionize Wearables — If They Have These Features
Meta's AI ambitions for Ray-Ban smart glasses: Meta's upcoming Connect 2024 event could position its Ray-Ban smart glasses as a leading AI hardware device, pending key feature announcements. The Ray-Ban smart glasses, a collaboration between Meta and EssilorLuxottica, have already gained attention for their stylish design and wearability. Meta's potential to leverage its vast user data and AI capabilities could transform these glasses from a novelty item to an essential AI-powered device. The success of the smart glasses hinges on Meta's ability to introduce innovative features that seamlessly integrate AI into users' daily lives. Five critical features for AI dominance:...
read Sep 24, 2024Meta to Add Celebrity Voices to Its AI Chatbot
Meta's AI chatbot gets a celebrity makeover: Meta, the parent company of Facebook, is set to introduce celebrity voices to its AI chatbot, signaling a continued interest in blending star power with artificial intelligence. Meta has reportedly secured deals with notable celebrities including Awkwafina, Kristen Bell, John Cena, Judi Dench, and Keegan-Michael Key to lend their voices to the company's AI chatbot. This move comes after earlier rumors of Meta seeking such partnerships, indicating a strategic shift in how the company approaches AI personalization. AI and celebrity collaborations: While Meta has moved away from celebrity-themed chatbots, the incorporation of famous...
read Sep 22, 2024What to Expect at Meta Connect 2024
Event details and format: Meta's annual developer conference, Meta Connect 2024, is set to take place on Wednesday, September 25, with a focus on the company's mixed reality and AI initiatives. The event kicks off at 10 am Pacific time with a one-hour keynote presentation by Meta CEO Mark Zuckerberg. Following Zuckerberg's address, Meta CTO Andrew Bosworth will lead a developer-focused session at 11 am. Viewers can tune in through various platforms, including the Meta Connect website, Meta's YouTube channel, or in virtual reality via Meta Horizon. Anticipated hardware announcements: While no groundbreaking high-end VR headset is expected, Meta may...
read Sep 21, 2024People Like Ray-Ban Smart Glasses — Can Meta Win the AI Hardware Race?
Meta's smart glasses lead AI wearables race: The Ray-Ban Meta smart glasses have emerged as a frontrunner in the AI wearables market, outperforming expectations in a year when other AI gadgets failed to deliver. The smart glasses, a collaboration between Meta and Ray-Ban, have succeeded where other AI wearables and smart glasses have struggled, surpassing even Meta's own projections. Unlike other AI-focused devices that overpromised and underdelivered, the Ray-Ban Meta smart glasses position AI as one feature among many, contributing to their success. The glasses offer a range of functionalities, including photo and video capture, audio playback, and AI-powered features...
read Sep 20, 2024Bugs in Meta’s AI Ad Tools are Costing Customers a Lot of Money
Meta's ad platform faces widespread issues: Media buyers report significant challenges with Meta's advertising tools, including targeting errors and budget overruns, potentially costing clients substantial amounts of money. Ongoing platform bugs and disruptions: Meta's ad platform has experienced numerous issues since Memorial Day, with some buyers encountering problems as frequently as every other week. A Meta-maintained website tracking Facebook Ads Manager issues has noted at least a dozen disruptions since Memorial Day, including four classified as "major disruptions." Media buyers report various problems, such as improper ad loading and daily budgets being depleted within hours. Some advertisers estimate losses in...
read Sep 13, 2024Meta Announces Plans to Train AI on UK User Data
Meta expands AI training data to include UK user content: The tech giant announces plans to incorporate public posts, comments, and photos from British Facebook and Instagram users into its AI training datasets. Meta aims to leverage this data to accelerate the development and deployment of its generative AI products in the UK market. The company states that this initiative will help its AI systems better reflect "British culture, history, and idiom," though the specifics of this goal remain unclear. The data collection process will commence in the coming months, affecting adult accounts on both Facebook and Instagram platforms. Data...
read Sep 12, 2024If You Didn’t Set Your Photos to Private, Meta Probably Trained Its AI on Them
Meta's extensive data usage for AI training: Meta has confirmed that it has been using public posts and photos from adult Facebook and Instagram users since 2007 to train its artificial intelligence models. The revelation came during an Australian government inquiry into AI adoption, where Meta's global privacy director, Melinda Claybaugh, initially denied but later confirmed the extent of data usage. This practice includes all public text and photo content posted by adult users on Facebook and Instagram over the past 17 years. Users who have not explicitly set their posts to private have had their data included in Meta's...
read Sep 12, 2024Facebook and Instagram Make AI labels Less Prominent on Edited Content
Meta revamps AI content labeling on social platforms: The company is updating how it identifies and labels AI-generated or AI-edited content across Facebook, Instagram, and Threads, aiming to provide more nuanced information about the extent of AI involvement in content creation. Key changes to AI labeling: Meta is moving its "AI Info" tag from a prominent position beneath user names to a more discreet menu in the top-right corner of images and videos. Users will now need to actively click on the menu to access information about potential AI-generated or AI-edited content. This change applies to content that has been...
read Sep 11, 2024Meta Mines 16 Years of Social Data for AI Training
Meta's extensive data mining for AI training revealed: Meta has admitted to scraping all public posts from Facebook and Instagram since 2007 to train its generative AI model, raising significant privacy concerns and contrasting sharply with Apple's approach to AI development. Scale of data collection: Meta's admission covers a vast trove of user-generated content spanning nearly two decades, encompassing billions of posts, photos, and comments from its social media platforms. The data collection includes all public posts made on Facebook and Instagram since 2007, excluding only content from users under 18 and private accounts. Photos of children posted by parents...
read Sep 11, 2024Meta Rolls Out AI Features for Quest Headsets
Meta Quest update brings quality-of-life improvements: Meta has rolled out a new update for its Quest 3, Quest 2, and Quest Pro virtual reality headsets, introducing several enhancements to user experience and functionality. The update, version 69, comes ahead of the upcoming Meta Connect conference scheduled for September 25-26, 2024. While not featuring any major overhauls, the update focuses on refining existing features and adding new capabilities to improve overall usability. Meta AI gets voice activation: Users in the US and Canada can now activate Meta AI using a voice command, streamlining access to the virtual assistant. The "Hey Meta"...
read Sep 11, 2024Facebook Admits to Scraping Australians’ Photos, Posts to Train AI
Facebook's massive data collection for AI training: The social media giant has confirmed it is scraping public data from all Australian adult users on its platform to train AI models, without offering an opt-out option. Facebook is collecting public photos, posts, and other data from Australian adult users' accounts dating back to 2007 for AI training purposes. The company initially denied this practice but later confirmed it when pressed during an inquiry. Data from users under 18 is not scraped, but public photos of children posted on adult accounts are included in the collection. Discrepancy in user privacy options: Facebook...
read Sep 4, 2024Meta Challenges OpenAI with Ambitious AI Assistant Push
AI assistant race heats up: Meta's CEO Mark Zuckerberg aims to make MetaAI the most used AI assistant by the end of the year, challenging OpenAI's dominance in the field. Meta has rapidly pushed its AI assistant to its 3 billion users since its debut, reaching 400 million monthly users and 40 million daily users earlier this month. OpenAI's ChatGPT currently leads with 200 million weekly active users, setting a high bar for competitors in the AI space. The aggressive promotion of MetaAI has raised concerns within the company about potential user fatigue and retention issues. User adoption challenges: Meta's...
read Sep 4, 2024Meta’s Llama Models Have Been Downloaded Nearly 350 Million Times This Year
Open-source AI surge led by Meta: Meta's Llama family of AI models has experienced explosive growth, with downloads approaching 350 million on Hugging Face, representing a tenfold increase from the previous year. Meta's mid-year update reveals significant adoption of its open-source AI models, particularly among large enterprises such as Zoom, Spotify, Infosys, AT&T, and Goldman Sachs. The performance of open-source AI is now matching that of closed-source alternatives, gaining substantial traction at the enterprise level. This trend is reshaping the AI landscape, putting pressure on closed-model companies to innovate and reduce costs. Meta's strategic open approach: The company's decision to...
read Aug 26, 2024Meta’s New AI System Slashes Root Cause Analysis Time
AI-driven root cause analysis revolutionizes incident response at Meta: Meta has developed an AI-assisted system that streamlines system reliability investigations, significantly reducing the time and effort required to identify root causes of incidents. Innovative approach combines heuristics and language models: The system utilizes a two-step process to efficiently narrow down potential root causes: A heuristic-based retriever first reduces the search space from thousands of code changes to a few hundred, using factors like code ownership and runtime code graphs. A large language model (LLM)-based ranker then analyzes these changes to identify the top five most likely root causes. Impressive accuracy...
read Aug 21, 2024Meta’s Self-Taught Evaluator Allows AI Models to Create Their Own Training Data
Breakthrough in AI training: Meta's researchers have developed a Self-Taught Evaluator, a groundbreaking approach that enables large language models (LLMs) to create their own training data without human annotation. This innovative method addresses the significant challenge of expensive and time-consuming human evaluation for LLMs, potentially revolutionizing the way AI models are trained and evaluated. The Self-Taught Evaluator builds upon the concept of LLM-as-a-Judge, where the model itself evaluates responses to given prompts, eliminating the need for human intervention. Meta's approach marks a significant step towards more efficient and scalable AI development, particularly beneficial for enterprises with vast amounts of unlabeled...
read Aug 12, 2024Meta Announces Llama 3.1 Impact Grants Program to Support Open Source AI
Meta's Llama 3.1 Impact Grants program aims to support innovative applications of open source AI, building on the success of last year's initiative to address global challenges in education, the environment, and open innovation. Expanding AI accessibility: Meta is committed to making AI benefits widely available through open source technology, seeking to extend opportunities to more communities worldwide. The company believes AI can significantly enhance human productivity, creativity, and quality of life while driving economic growth and advancements in various fields. By offering open source AI models, Meta aims to democratize access to cutting-edge technology for individuals and organizations globally....
read Aug 12, 2024AI-Generated Images Flood Facebook, Exposing Moderation Flaws
The proliferation of AI-generated images on Facebook has led to an investigation uncovering a profitable marketplace for creating and sharing dubious content, raising questions about the platform's content moderation and reward systems. The rise of AI slop: Facebook is experiencing an influx of AI-generated images featuring bizarre and often nonsensical content, which are gaining significant traction on the platform. Images such as "shrimp Jesus," trucks full of babies, and police officers holding giant bibles in floodwaters have gone viral, accumulating tens of thousands of likes. These images, collectively referred to as "slop," represent a growing trend of AI-generated content flooding...
read Aug 11, 2024Meta and Oxford Develop AI Model that Turns Images to 3D Objects
VFusion3D, a groundbreaking AI model developed by Meta and the University of Oxford, represents a significant advancement in AI-powered 3D content creation, enabling the generation of high-quality 3D objects from single images or text descriptions. Innovative approach to 3D generation: VFusion3D overcomes the challenge of limited 3D training data by leveraging pre-trained video AI models to generate synthetic 3D data. This novel approach allows the model to create high-quality 3D objects without relying on extensive 3D datasets, which have been a bottleneck in previous attempts at AI-driven 3D generation. The use of video AI models for 3D data synthesis demonstrates...
read Aug 10, 2024Instagram’s AI Studio Fuels Surge in Virtual Companions
A new frontier in social media: Meta's recently launched AI Studio for Instagram has opened the doors to user-created AI chatbots, leading to an influx of virtual "boyfriends" and "girlfriends" on the platform. Meta's AI Studio allows US-based Instagram users to create digital versions of themselves or AI characters capable of conversing with others. The feature has quickly been adopted by users to create AI romantic partners, filling a niche for those seeking virtual companionship. This development marks a significant shift in how AI is integrated into mainstream social media platforms. The rise of AI companions: The emergence of AI...
read Aug 8, 2024What to Expect from Meta Connect 2024
Meta Connect 2024 announced: Meta has set the date for its annual developer conference, Meta Connect 2024, which will take place on September 25-26 at the company's headquarters in Menlo Park, California. Event format and focus: The two-day conference will showcase Meta's latest innovations in artificial intelligence, virtual reality, and the metaverse, offering both in-person and virtual attendance options. Meta will host the event in a hybrid format, allowing attendees to participate either on-site or virtually through Facebook and Horizon Worlds. The conference aims to provide developers, creators, and tech enthusiasts with insights into Meta's latest advancements and future plans....
read Aug 8, 2024Meta Bids Big on Celebrity Voices for AI Assistant
Meta's ambitious AI voice project: Meta is reportedly offering substantial sums to celebrities for the rights to use their voices in AI projects, with a particular focus on its digital assistant MetaAI. The tech giant is said to be in negotiations with major Hollywood talent agencies to secure voice rights from various high-profile celebrities, including Awkwafina, Judi Dench, and Keegan-Michael Key. Meta's goal is to obtain these voice rights before its Connect 2024 event in September, allowing time to develop AI tools incorporating the celebrity voices for showcase at the event. While the exact intended use of these voice rights...
read Aug 5, 2024Meta Unveils AI Network Architecture to Power Next-Gen Models
Meta's AI infrastructure revolution: Meta has developed specialized data center networks designed to support large-scale distributed AI training using GPU clusters, marking a significant advancement in AI infrastructure. The company's approach employs RDMA Over Converged Ethernet version 2 (RoCEv2) as the inter-node communication transport, highlighting the importance of high-speed, low-latency networking in AI workloads. Meta's network architecture is divided into two distinct parts: a frontend network for data ingestion, checkpointing, and logging, and a backend network specifically optimized for AI training tasks. AI Zone: The backbone of Meta's AI network: The backend network utilizes a two-stage Clos topology, dubbed an...
read