2
respostas

[Dúvida] Avaliando o método 70-20-10 com Claude AI

**Escrevi esse prompt para o Claude AI como alternativa ao GPT a fim de realizar uma atividade para o curso de aprendizagem "Aprendizagem Autônoma":

"Recently i knew about the learning method "70-20-10". And now I want to use it to explore my learning process in IT field, bacause I'm starting studying coding and other stuffs related to this field from scratch. I'm doing about 5 to 6 hours of formal studying per day, I'm having zero interactions with real people from the IT field and I'm not working with IT yet, only trying to apply my knowledge in small chalenges. Can you give me data using the 70-20-10? Saying and pointing where can I get better and where to do my adjusts?"

Eis a resposta: Let me help you analyze your current learning approach using the 70-20-10 model, which suggests that optimal learning comes from:

  • 70% hands-on experience and challenging assignments
  • 20% social learning and mentorship
  • 10% formal education and structured learning Looking at your current situation, you're spending 5-6 hours on formal learning (the 10% portion) but have minimal exposure to the other components. Let's break this down and suggest some adjustments: Your Current Distribution: You seem to be heavily weighted toward formal learning (your 5-6 hours of studying) with some practical application (your small challenges), but missing the social learning aspect entirely. This creates an imbalance that could slow down your overall progress. Here's how we can optimize each component: For the 70% (Experiential Learning): Instead of just small challenges, try building more substantial projects that mirror real-world applications. For example, if you're learning web development, create a complete website that solves a real problem. This helps transform your theoretical knowledge into practical skills. Consider dedicating about 7-8 hours daily to hands-on coding, troubleshooting, and building. For the 20% (Social Learning): This is currently your biggest gap. Here are several ways to address it:
  • Join coding communities on platforms like Discord or Reddit
  • Participate in local tech meetups (many cities have regular gatherings for developers)
  • Contribute to open-source projects on GitHub where you can interact with other developers
  • Find a study group or coding buddy online
  • Attend virtual tech conferences or workshops
  • Join coding-focused social media groups Aim to spend about 2-3 hours daily engaging with the community in some form. For the 10% (Formal Learning): Your current 5-6 hours of formal studying actually exceeds what's needed according to the model. Consider reducing this to about 1-2 hours daily of structured learning (courses, tutorials, documentation reading) and redirecting that time to hands-on practice and social learning. A Balanced Daily Schedule Might Look Like This:
  1. Morning: 1-2 hours of formal learning (watching courses, reading documentation)
  2. Mid-day: 7-8 hours of hands-on coding and project work
  3. Evening: 2-3 hours of community engagement (participating in forums, contributing to open source, attending virtual meetups) Remember that these numbers are flexible and should be adjusted based on your schedule and energy levels. The key is maintaining the approximate ratio over time, not necessarily every single day. To track your progress with this method, consider keeping a learning journal where you log:
  • Time spent on each type of learning activity
  • Skills gained from each experience
  • Challenges faced and overcome
  • Connections made with other developers
  • Projects completed This will help you ensure you're maintaining a healthy balance across all three learning components and make adjustments as needed. Would you like me to suggest some specific platforms or communities where you could start building your social learning network?

**A resposta que ele me forneceu foi bem mais satisfatória e organizada que o GPT e Gemini de maneira geral.

**Seria o sistema do Claude mais "preparado" para responder questões mais analíticas do que os outros dois modelos?

2 respostas

Oi Vinicius,

A análise do Claude AI sobre o método 70-20-10 foi bem detalhada e organizada, 👍 o que sugere que ele pode ter um bom desempenho em questões analíticas.

A forma como ele desmembrou cada componente do método e propôs ajustes práticos é um ponto forte. 🤔

É possível que o Claude tenha sido treinado com um foco maior em raciocínio e análise, o que o torna mais eficaz em tarefas que exigem esse tipo de abordagem.

No entanto, cada modelo de IA tem seus pontos fortes e fracos, e a escolha ideal pode variar dependendo da tarefa específica. 🧐

A sugestão de um cronograma diário e o acompanhamento do progresso em um diário de aprendizagem são excelentes para manter o equilíbrio entre as diferentes formas de aprendizado. 💪

Para saber mais: Artigo sobre o método 70-20-10.

70-20-10

Continue explorando e adaptando as ferramentas de IA para otimizar seu aprendizado! 💪

Muito obrigado pela resposta, Luis!

Agradeço também pelo artigo, vou conferir.