Tag: llm
-
Transformer: The quiet revolution that changed Artificial Intelligence forever
In the summer of 2017, a seemingly modest research paper titled “Attention Is All You Need” quietly emerged from Google Brain, fundamentally transforming the landscape of artificial intelligence. While it didn’t arrive with fanfare, this paper would become the foundation for virtually every major AI model we use today, from OpenAI’s ChatGPT to Meta’s Llama…
-
DeepSeek unveils V3 Language Model with remarkable efficiency
DeepSeek has introduced its latest advancement in artificial intelligence, the DeepSeek-V3, a revolutionary language model that combines exceptional performance with remarkable efficiency. This innovative system employs a Mixture-of-Experts (MoE) architecture, featuring 671 billion total parameters while activating only 37 billion for each token processing task. What sets DeepSeek-V3 apart is its unprecedented training efficiency. The…
-
OpenAI faces second major December outage: Make local AI processing more appealing
OpenAI experienced another significant service disruption on Thursday, with ChatGPT, Sora, and its developer APIs going dark for over four hours, marking the second major outage this month. The incident, which began at 11 a.m. PT, affected millions of users worldwide and has reignited discussions about the reliability of cloud-based AI services. The company attributed…
-
UAE’s TII launches Falcon 3: A new generation of efficient Language Models
The Technology Innovation Institute (TII), backed by the UAE government, has introduced Falcon 3, a significant advancement in Small Language Model (SLM) technology. This new family of open-source models represents a strategic move toward more accessible and efficient AI implementations. The Falcon 3 series comprises four model variants—1B, 3B, 7B, and 10B parameters—each available in…
-
Connecting data: Perplexity acquires Carbon to enhance AI search
Perplexity AI has made a significant move in the tech landscape by acquiring Carbon, a Seattle-based startup specializing in data connectivity for large language models. This strategic acquisition aims to enhance Perplexity’s AI capabilities by integrating Carbon’s advanced retrieval engine, which connects external data sources to AI systems. Users can expect to link popular applications…
-
xAI expands Grok’s accessibility with new iOS app and web platform
xAI, Elon Musk’s artificial intelligence venture, is broadening access to its Grok chatbot through a new iOS application, currently in beta testing across select countries including Australia. This expansion marks a significant shift from Grok’s previous exclusivity to X (formerly Twitter) platform users. The standalone application showcases comprehensive AI capabilities, incorporating real-time data access from…
-
Top open-source Language Models in 2024
The landscape of open-source Language Models continues to evolve, with various models offering unique capabilities. Here’s a comprehensive overview of the top models currently shaping the industry in 2024. LLaMA 3 (8B-70B parameters) Meta’s upgraded model offers two variants: 8B and 70B parameters. The 70B version demonstrates exceptional efficiency in language modeling and question-answering tasks.…
-
OpenAI expands ChatGPT’s web search capabilities to more users
OpenAI is broadening the reach of its AI-powered web search tool within ChatGPT, extending access to a larger user base. This move significantly enhances ChatGPT’s capabilities, transforming it from a purely language-based model into a more versatile tool capable of accessing and integrating real-time information from the web. Previously, the web browsing feature was limited…
-
Meta launches Llama 3.3: A compact powerhouse in open-source AI
Meta has launched Llama 3.3, an open-source language model with 70 billion parameters that matches its predecessor’s performance while reducing computational costs. It supports multiple languages and offers a large token context window. The model emphasizes efficiency, responsible AI development, and environmental sustainability, available for download from various platforms.
-
Microsoft introduces Phi-4 language model for Complex Reasoning: Smaller, Smarter
Microsoft has launched Phi-4, its latest Small Language Model (SLM), designed to excel in complex reasoning and language tasks while maintaining impressive efficiency. Building on the success of previous Phi models, Phi-4 aims to push the boundaries of what’s possible with smaller, more accessible AI. Unlike Large Language Models (LLMs) which are often computationally intensive,…