Carlos Monteiro

1 month ago | expresso.pt | Tiago Oliveira |Carlos Monteiro

Em resposta à pergunta do título: é a força de apoio contínuo das cerca de 74 mil entidades que compõem o sector social e cujos trabalhadores representam — segundo a Cooperativa António Sérgio para a Economia Social (CASES) e o Instituto Nacional de Estatística — perto de 5,9% do emprego total, e 3,2% do VAB (Valor Acrescentado Bruto) português, o que equivale a mais de €5,5 mil milhões, de acordo com os últimos dados.

1 month ago | medium.com | Carlos Monteiro

Member-only storyCarlos Monteiro·Follow13 min read·--Running your own large-scale language model locally was once a daunting challenge. But with Gemma 3, Google’s advanced, multimodal LLM, you can enjoy powerful AI capabilities on personal hardware — securely, privately, and (often) affordably. This guide covers everything you need to know to get Gemma 3 up and running, from basic setup and performance tuning to integrating with AI agent frameworks like LangChain or AutoGPT.

1 month ago | medium.com | Carlos Monteiro

When Google rolls out a new large language model, the AI community pays attention. Meet Gemma 3: Google’s boldest, most ambitious entry yet into the booming world of open-source large language models. Built from the groundbreaking foundations of Google’s Gemini 2.0, Gemma 3 is designed not just to impress — but to revolutionize what you can achieve locally, right in your own home.

1 month ago | medium.com | Carlos Monteiro

Member-only storyCarlos Monteiro·Follow8 min read·--In the rapidly evolving landscape of artificial intelligence, Retrieval-Augmented Generation (RAG) stands out as a transformative approach that enhances the capabilities of Large Language Models (LLMs). By integrating a retrieval mechanism with generative power, RAG enables these models to tap into external knowledge sources, ensuring responses that are not only coherent but also factually grounded.

1 month ago | medium.com | Carlos Monteiro

Member-only storyCarlos Monteiro·Follow6 min read·--(Image: OpenAI)On February 27, 2025, OpenAI introduced GPT-4.5, a monumental advancement in large language models that has quickly captured the attention of researchers, developers, and AI enthusiasts alike. Hailed as the company’s largest and most knowledgeable model to date, GPT-4.5 builds upon the foundation of its predecessor, GPT-4o, promising greater accuracy, fewer hallucinations, and a conversational finesse that feels remarkably human.