NewApp Digest
search bot

Llama

Meta утверждает, что раздача пиратских книг через торренты подпадает под определение "добросовестного использования"
TrendTechie
March 12, 2026

Meta утверждает, что раздача пиратских книг через торренты подпадает под определение “добросовестного использования”

Meta is facing a class-action lawsuit from authors over the use of pirated books for training its Llama language model, arguing that the distribution of these books via BitTorrent constitutes fair use. The company used shadow libraries like Anna's Archive to gather text, and a California court has partially ruled that using pirated books for training large language models falls under fair use, although the issue of copyright infringement related to downloading and distribution via BitTorrent remains unresolved. The plaintiffs claim Meta has not previously indicated a fair use defense regarding file-sharing, while Meta asserts that authors have not shown evidence of their works being reproduced by the Llama model. The judge will ultimately decide on the admissibility of Meta's defense.
Mob Menagerie: Llama
AppWizard
November 15, 2025

Mob Menagerie: Llama

Llamas thrive in diverse environments, including savanna plateaus, windswept hills, gravelly terrains, and forest biomes. In the Bedrock Edition, they can also be found in standard savanna regions. A typical group of llamas consists of four to six individuals, displaying colors such as rich browns, creamy whites, and soft grays. They are generally indifferent to human presence but may spit if approached too closely. Llamas can defend themselves against wild wolves but are tolerant of tamed wolves if not provoked. Taming a llama requires persistence, involving mounting and offering wheat. Once tamed, llamas can carry a chest with three to fifteen slots and can be adorned with carpets for unique designs. Llamas can form a caravan by leashing one, allowing for the transportation of goods over long distances.
AI-торрент: децентрализованная сеть, где твои GPU станут валютой для супер-AI
TrendTechie
November 12, 2025

AI-торрент: децентрализованная сеть, где твои GPU станут валютой для супер-AI

A decentralized P2P network called "AI Torrent" is designed for AI model inference, based on principles such as a BitTorrent economy where nodes exchange computational resources, a Smart Swarm Architecture using specialized models, and Self-Organizing Intelligence that allows popular models to migrate to active nodes. Users without computational resources can interact through a standard API or chat interface, while those with resources can register as nodes and earn utility tokens by performing inference tasks. AI model creators can upload models and receive royalties through smart contracts. The economy operates on utility tokens (AIT), with revenue distribution of 70% to seeders, 20% to model developers, and 10% to a DAO fund. The network aims to be cheaper than centralized alternatives and has mechanisms for liquidity and stability, including trading on DEX platforms and staking. Existing projects in decentralized AI demonstrate the viability of P2P inference, and the "AI Torrent" seeks to integrate their best features while focusing on making inference accessible and affordable. Challenges include latency in P2P systems, which the network aims to address through geo-DHT and edge caching.
Dreaming in Blocks — MineWorld, the Minecraft World Model
AppWizard
October 10, 2025

Dreaming in Blocks — MineWorld, the Minecraft World Model

Microsoft has unveiled Mineworld, an open-source world model for enhancing the Minecraft experience, which allows for real-time interactions and high controllability. Mineworld improves upon previous models, particularly the closed-source Oasis, by addressing computational inefficiencies. Its key components include: - Real-time interactivity for fast, dynamic gameplay. - A parallel decoding algorithm that accelerates frame generation. - A novel evaluation metric for assessing controllability. Mineworld processes video game footage and player actions through unique tokenization methods. Each frame in a video clip consists of quantized tokens representing distinct pixel sets. Player actions are tokenized into 11 distinct tokens, including 7 for action groups, 2 for camera angles, and 2 for action sequence markers. Mineworld was released on April 11, 2025.
18 of the best Minecraft tips and secrets to help you become a master builder
AppWizard
August 8, 2025

18 of the best Minecraft tips and secrets to help you become a master builder

Minecraft has 204.33 million monthly players across mobile devices and consoles. Players can use llamas as storage units, break blocks for navigation, utilize lava as a fuel source, and find diamonds most effectively between levels 15 and 63. Players can change sheep colors using Evokers, harvest sand or gravel with torches, and explore Mushroom biomes safely. Strategies include wearing carved pumpkins to avoid Endermen attacks, creating Iron Golems for protection, using cats to deter Creepers, and placing doors underwater for air pockets. Snowballs are effective against Blazes, and torches can stop mob spawners. Players can milk Mooshrooms for stew, customize dog collars, protect pixel art with glass panes, and lure zombies with turtle eggs.
18 essential Minecraft tip and tricks that will help you master the game
AppWizard
August 8, 2025

18 essential Minecraft tip and tricks that will help you master the game

Minecraft has approximately 204.33 million monthly players and is available on all mobile devices and gaming consoles. Llamas can be used as pack mules, with inventory capacity varying from 3 to 15 slots based on strength. Players can break blocks to leave a breaking pattern on the north side for navigation. Lava is a superior fuel source compared to wood or coal. Diamonds are now typically found between levels 15 and 63. Blue sheep can be turned red by an Evoker. Torches can help harvest sand and gravel more effectively. Mushroom biomes do not spawn aggressive mobs at night. Carved pumpkins can prevent Endermen attacks. Iron Golems can be created for protection using iron blocks and a carved pumpkin. Cats deter Creepers. Doors create air pockets underwater. Snowballs are effective against Blazes. Placing torches around mob spawners can halt enemy production. The Swift Sneak enchantment allows quick movement while crouching. Mooshrooms produce mushroom stew when milked. Players can dye their dogs' collars for identification. Adding a glass pane to a map preserves pixel art. Zombies are attracted to turtle eggs, which can be used to lure them away.
AWS Offers OpenAI’s Models on Its Platform for the First Time
Winsage
August 6, 2025

AWS Offers OpenAI’s Models on Its Platform for the First Time

OpenAI has made its gpt-oss models, with 120 billion and 20 billion parameters, available for public use on AWS's Bedrock platform, marking the first open model release since GPT-2 in 2019. These models are not entirely open source, as users do not have access to the underlying code or training datasets. AWS's chief evangelist, Danilo Poccia, expressed enthusiasm for the integration of these models, which are designed for coding, scientific analysis, and mathematical reasoning, and can be integrated into business workflows. Analysts note that OpenAI's models provide substantial value compared to competitors, aligning with Amazon's cost-saving strategy. AWS hosts various open models, but Claude from Anthropic was absent from the recent press release, indicating a complex competitive landscape. The addition of OpenAI's models to AWS is seen as a potential step towards deeper collaboration between the two companies.
OpenAI’s New Open Models Accelerated Locally on NVIDIA GeForce RTX and RTX PRO GPUs
Winsage
August 6, 2025

OpenAI’s New Open Models Accelerated Locally on NVIDIA GeForce RTX and RTX PRO GPUs

NVIDIA has partnered with OpenAI to enhance the gpt-oss models for NVIDIA GPUs, enabling rapid inference and supporting millions of users on NVIDIA RTX AI PCs. The gpt-oss-20b and gpt-oss-120b models, trained on NVIDIA H100 GPUs, feature open-weight reasoning and can handle context lengths of up to 131,072 tokens. Users can utilize these models through frameworks like Ollama, which provides a user-friendly interface for experimentation. The models are optimized for RTX GPUs and support applications such as web search and coding assistance. Developers can also access the models via Microsoft AI Foundry Local and other frameworks, with NVIDIA contributing to open-source projects to enhance performance.
Search