<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="pt-BR">
  <title>Codeflow</title>
  <subtitle>Artigos e Referências de Tecnologia</subtitle>
  <id>https://codeflow.com.br/pt-BR/</id>
  <link rel="alternate" type="text/html" href="https://codeflow.com.br/pt-BR/"/>
  <link rel="self" type="application/atom+xml" href="https://codeflow.com.br/pt-BR/feed.xml"/>
  <updated>2026-09-10T16:00:00-03:00</updated>
  <author><name>Rafael Santos</name><uri>https://github.com/rafaelsantos</uri></author>
  <icon>https://codeflow.com.br/assets/icons/icon-192.png</icon>
  <rights>Codeflow © 2026</rights>
  <entry>
    <title>Mitologia, autômatos e a origem da pergunta: o início da história da IA, de Turing ao primeiro inverno</title>
    <id>https://codeflow.com.br/pt-BR/articles/mythology-automata-and-the-origin-of-the-question-the-early-history-of-ai/</id>
    <link rel="alternate" type="text/html" href="https://codeflow.com.br/pt-BR/articles/mythology-automata-and-the-origin-of-the-question-the-early-history-of-ai/"/>
    <published>2026-09-10T16:00:00-03:00</published>
    <updated>2026-09-10T16:00:00-03:00</updated>
    <author><name>Rafael Santos</name></author>
    <category term="Artificial Intelligence" label="Inteligência Artificial"/>
    <category term="history-of-ai"/>
    <category term="alan-turing"/>
    <category term="turing-test"/>
    <category term="dartmouth-conference"/>
    <category term="perceptron"/>
    <category term="neural-networks"/>
    <category term="expert-systems"/>
    <category term="decision-trees"/>
    <category term="ai-winter"/>
    <category term="symbolic-ai"/>
    <summary type="text">O início da história da IA: mitos e autômatos, a pergunta e o teste de Turing, Dartmouth, perceptrons, sistemas de regras e o caminho até o primeiro inverno.</summary>
  </entry>
  <entry>
    <title>Da previsão de saída ao mapa probabilístico: autorregressão, self-attention e embeddings</title>
    <id>https://codeflow.com.br/pt-BR/articles/from-output-prediction-to-the-probabilistic-map-autoregression-self-attention-and-embeddings/</id>
    <link rel="alternate" type="text/html" href="https://codeflow.com.br/pt-BR/articles/from-output-prediction-to-the-probabilistic-map-autoregression-self-attention-and-embeddings/"/>
    <published>2026-09-10T14:00:00-03:00</published>
    <updated>2026-09-10T14:00:00-03:00</updated>
    <author><name>Rafael Santos</name></author>
    <category term="Artificial Intelligence" label="Inteligência Artificial"/>
    <category term="generative-ai"/>
    <category term="llm"/>
    <category term="autoregressive-models"/>
    <category term="self-attention"/>
    <category term="embeddings"/>
    <category term="transformers"/>
    <category term="next-token-prediction"/>
    <category term="semantic-similarity"/>
    <category term="hallucination"/>
    <category term="deep-learning"/>
    <summary type="text">Como os modelos generativos compõem: um mapa probabilístico em vez de um mapeamento, autorregressão token a token, self-attention e embeddings.</summary>
  </entry>
  <entry>
    <title>IA tradicional versus IA generativa: prever a partir de dados conhecidos ou compor novas saídas</title>
    <id>https://codeflow.com.br/pt-BR/articles/traditional-ai-versus-generative-ai-predict-from-known-data-or-compose-new-outputs/</id>
    <link rel="alternate" type="text/html" href="https://codeflow.com.br/pt-BR/articles/traditional-ai-versus-generative-ai-predict-from-known-data-or-compose-new-outputs/"/>
    <published>2026-09-10T08:00:00-03:00</published>
    <updated>2026-09-10T08:00:00-03:00</updated>
    <author><name>Rafael Santos</name></author>
    <category term="Artificial Intelligence" label="Inteligência Artificial"/>
    <category term="traditional-ai"/>
    <category term="generative-ai"/>
    <category term="supervised-learning"/>
    <category term="classification"/>
    <category term="regression"/>
    <category term="labeled-data"/>
    <category term="fraud-detection"/>
    <category term="self-supervised-learning"/>
    <category term="transformers"/>
    <category term="llm"/>
    <summary type="text">A IA tradicional prevê uma classe ou um valor dentro de um espaço de respostas fechado; a IA generativa compõe uma saída nova a partir de padrões aprendidos.</summary>
  </entry>
  <entry>
    <title>Quando o código deixa de ser o diferencial: o papel do desenvolvedor com IA generativa</title>
    <id>https://codeflow.com.br/pt-BR/articles/when-code-stops-being-the-differentiator-the-developers-role-with-generative-ai/</id>
    <link rel="alternate" type="text/html" href="https://codeflow.com.br/pt-BR/articles/when-code-stops-being-the-differentiator-the-developers-role-with-generative-ai/"/>
    <published>2026-09-09T14:00:00-03:00</published>
    <updated>2026-09-09T14:00:00-03:00</updated>
    <author><name>Rafael Santos</name></author>
    <category term="Artificial Intelligence" label="Inteligência Artificial"/>
    <category term="generative-ai"/>
    <category term="developer-role"/>
    <category term="productivity"/>
    <category term="technical-leadership"/>
    <category term="prompt-engineering"/>
    <category term="ai-agents"/>
    <category term="continuous-learning"/>
    <category term="software-engineering"/>
    <category term="human-in-the-loop"/>
    <category term="llm"/>
    <summary type="text">Por que escrever código já não basta: para onde migra o diferencial do desenvolvedor, como a IA amplia competência e o que muda para liderança e produto.</summary>
  </entry>
  <entry>
    <title>Fundamentos antes de ferramentas: como avaliar e construir com IA generativa</title>
    <id>https://codeflow.com.br/pt-BR/articles/fundamentals-over-tools-how-to-evaluate-and-build-with-generative-ai/</id>
    <link rel="alternate" type="text/html" href="https://codeflow.com.br/pt-BR/articles/fundamentals-over-tools-how-to-evaluate-and-build-with-generative-ai/"/>
    <published>2026-09-09T08:00:00-03:00</published>
    <updated>2026-09-09T08:00:00-03:00</updated>
    <author><name>Rafael Santos</name></author>
    <category term="Artificial Intelligence" label="Inteligência Artificial"/>
    <category term="generative-ai"/>
    <category term="llm"/>
    <category term="machine-learning"/>
    <category term="deep-learning"/>
    <category term="model-selection"/>
    <category term="trade-offs"/>
    <category term="llm-engineering"/>
    <category term="rag"/>
    <category term="fine-tuning"/>
    <category term="latency"/>
    <summary type="text">Por que entender os fundamentos da IA generativa vale mais do que dominar ferramentas: camadas, mecanismo, critérios de engenharia e o modelo como componente.</summary>
  </entry>
  <entry>
    <title>O que é IA generativa, e por que ela só se tornou viável agora?</title>
    <id>https://codeflow.com.br/pt-BR/articles/what-is-generative-ai-and-why-did-it-only-become-viable-now/</id>
    <link rel="alternate" type="text/html" href="https://codeflow.com.br/pt-BR/articles/what-is-generative-ai-and-why-did-it-only-become-viable-now/"/>
    <published>2026-09-08T09:00:00-03:00</published>
    <updated>2026-09-08T09:00:00-03:00</updated>
    <author><name>Rafael Santos</name></author>
    <category term="Artificial Intelligence" label="Inteligência Artificial"/>
    <category term="generative-ai"/>
    <category term="llm"/>
    <category term="transformers"/>
    <category term="prompt-engineering"/>
    <category term="deep-learning"/>
    <category term="attention"/>
    <category term="gpu"/>
    <category term="human-in-the-loop"/>
    <summary type="text">Uma introdução à IA generativa: os três fatores que a tornaram viável, a diferença para a IA tradicional, o prompt como controle e a revisão humana obrigatória.</summary>
  </entry>
</feed>
