Google releases new AI model Gemini, but admits video demonstration is fabricated
2023-12-08
Google accused of overstating performance of Gemini AI model in a video, with claims of distorted capabilities. Video disclaimer mentions reduced latency and edited output. Critics argue real-time interaction misrepresented. Google defends video as shortened for brevity, created to inspire developers. Concerns raised about misleading flashy videos overshadowing AI capabilities.
Read more ›Inflection AI Releases Android Version Pi: An AI Chatbot App Supporting Voice Interaction
2023-12-11
Inflection AI launched AI chatbot "Pi" as a dedicated Android app, aiming to rival ChatGPT and other AI assistants. Pi offers voice interactions, improved branding, and plans for enhanced capabilities with Inflection-2, emphasizing emotional support and personalized conversations.
Read more ›EU adopts world's first comprehensive artificial intelligence regulatory agreement
2023-12-11
EU reaches historic agreement on "Artificial Intelligence Act," regulating generative AI and facial recognition. Groundbreaking step in global tech governance, setting standards for AI regulations. Critics cite concerns over exemptions, privacy infringements, and potential loopholes.
Read more ›Apple provides developers with the MLX framework for machine learning.
2023-12-11
Apple released MLX, an array framework for building ML transformer models on its silicon. MLX aims to be user-friendly with easy deployment and efficient model training. Apple is catching up in the generative AI field with plans to launch AI features for iOS and Siri.
Read more ›Mistral AI releases new open-source large model Mixtral MoE 8x7B
2023-12-11
Open-source Mistral AI releases new LLM seed link, contrasting with Google's Gemini. Mistral's MoE 8x7B model described as scaled-down version of GPT-4 with Mixture of Experts. AI community praises Mistral's unconventional release method. Startup valued at $2 billion.
Read more ›Google Launches "Project Ellman": Using AI to Tell Personal Life Stories
2023-12-11
A team at Google introduced "Project Ellmann," utilizing AI to create a "bird's-eye view" of users' lives. This innovative approach aims to enhance Google Photos and other products. Google also launched Gemini, an advanced AI model with multimodal capabilities. Apple is also exploring personalized memory features.
Read more ›The Dominance of Large Tech Companies Stifles Artificial Intelligence Innovation
2023-12-11
Tech startups rely on giants like Microsoft, Amazon, Google for AI infrastructure & market access. Small companies mirror giants' AI models; often get acquired. Dominant tech firms collect user data, shape AI future, limit competition. Balance needed for ethical AI progress. Dominance by big tech disrupts market, reduces innovation, privacy, autonomy. Interventions crucial to prevent further consolidation of power.
Read more ›Google Upgrades Bard: Introducing Gemini Model, Performance Surpasses Across the Board
2023-12-07
OpenAI's ChatGPT is popular, surpassing Google's Bard. But Bard now runs on Gemini model, claiming to rival ChatGPT with new features and better technology. Google plans to expand Gemini to more languages. Gemini Pro aims to be fast and efficient. Next year, Bard Advanced will launch, powered by Gemini Ultra, a multimodal model. Gemini allows non-text interaction for tasks like coding. Google believes Gemini will make Bard as good as ChatGPT.
Read more ›Apple Inc. Releases Model Framework, Joins AI Competition.
2023-12-07
Apple's ML research team released MLX, a framework for efficient model building on Apple Silicon. It integrates with MLX Data deep learning models and popular frameworks like PyTorch, focusing on flexibility and ease of use. Apple aims to optimize AI applications on its devices, emphasizing machine learning over generative AI.
Read more ›Google releases Gemini Nano: a large language model that can run on mobile phones
2023-12-07
Google introduces Gemini Nano, a smaller version of Gemini LLM for mobile devices. Pixel 8 Pro users can now enjoy features like automatic summarization and Smart Reply offline. More applications and Bard integration coming next year. Nano aims to maximize performance without storage or heating issues. AICore system service allows integration with compatible processors. Gemini Nano enhances Pixel phones' AI capabilities.
Read more ›Mistral AI is set to secure another $487 million in funding.
2023-12-07
The founders of Mistral AI agree to sell shares, securing record seed funding and raising €450 million. Investors include NVIDIA, Salesforce, and Andreessen Horowitz. Mistral aims to provide cost-effective AI solutions and is valued at $2 billion.
Read more ›"Musk's AI Startup Plans to Raise $1 Billion"
2023-12-07
Elon Musk's xAI seeks $1B funding to rival OpenAI's ChatGPT. Musk's company raised $134.7M, plans full $1B raise, unveils ChatGPT-like chatbot "Grok." xAI recruits researchers, aims to understand the universe. Tech giants compete in generative AI field, OpenAI delays stock sale.
Read more ›EU Resumes Negotiations on World's First AI Law
2023-12-08
Negotiations on EU's comprehensive AI law lasted 22 hours. ChatGPT's rapid capabilities highlighted the urgency. Challenges include regulating foundational AI models and remote biometric surveillance. EU aims for 2023 approval with possible delay until 2026.
Read more ›AMD's ROCm Ready to Challenge NVIDIA's CUDA
2023-12-08
AMD unveils the MI300X, a faster alternative to NVIDIA GPUs, alongside open-source ROCm to challenge CUDA. Partnerships with AI startups highlight performance benefits and strategic investments in AI software enhance capabilities. ROCm 6 on MI300X outperforms MI250X and CUDA, supporting new data types and attention algorithms for improved AI performance.
Read more ›Elon Musk's new AI "Grok" exclusively lands on X social platform, challenging ChatGPT.
2023-12-08
ChatGPT competitor Grok launched on X offering ad-free social access for $16/mo. Grok answers in conversational manner, pulls real-time data from X. Grok can be cheeky, using profanity and being colloquial. Limited to text input for now, but plans to expand to handle videos and audio. Musk focuses on subscription services amid advertiser exits. X to introduce paid services like peer-to-peer payments.
Read more ›Meta Launches Purple Llama Initiative: Empowering Developers to Build Secure and Reliable Generative AI
2023-12-08
Meta launches Purple Llama initiative to empower developers in responsible AI, offering CyberSec Eval and Llama Guard tools. Over 100 partners join to address AI challenges. Purple symbolizes comprehensive cybersecurity strategy. Collaboration key in advancing ethical AI.
Read more ›Lutra AI, a startup, launches a platform that simplifies building AI workflows.
2023-12-08
Lutra AI, led by ex-Google and Coursera employees, raised $3.8M for customizable AI workflows. It focuses on code-first integration using advanced natural language AI models to create reliable and secure workflows connecting tools like Gmail and Slack. The startup targets non-technical users to automate tasks, emphasizing productivity and efficiency. The platform is in private testing, with support from Coatue Ventures, Hustle Fund, and Maven Ventures.
Read more ›Microsoft Bing Launches Deep Search: A New Feature Expanding Queries with GPT-4
2023-12-06
Microsoft is introducing a new Deep Search feature for Bing, powered by GPT-4. It provides detailed answers to vague queries, including loyalty card details and comparisons, immigration policies, and more. Currently in experimental stage for select users.
Read more ›Why Amazon Q Deserves Another Chance
2023-12-06
At AWS re:Invent, Amazon Q was introduced as a safer AI chatbot compared to OpenAI's ChatGPT. However, concerns about accuracy and privacy have arisen due to leaked sensitive data. Despite criticism, Amazon Q offers unique integration with popular systems. Other companies like OpenAI are also enhancing their chatbots. Despite challenges, AI chatbots are the future.
Read more ›DemoFusion: Democratizing High-Resolution Image Generation
2023-12-06
In the AI image generation field, DemoFusion democratizes high-resolution image creation, requiring only a consumer-grade RTX 3090 GPU. By leveraging the SDXL model, it enables users to generate high-quality images without extensive training, providing a more accessible alternative to centralized services.
Read more ›