You’re just starting out in AI. What are the most important skills to learn?

Python is the way to start, no frameworks, go with the fundamentals, move to data manipulation with Pandas and Numpy and proceed to learn Scikit-Learn. After you are comfortable with those, you can work in more complex applications using Keras, TensorFlow (and some other frameworks). In case you want to be ahead of the curve, you might want to take a look at MOJO. Which combines the usability of Python with better performance and might be the next frontier for AI Modeling.

As a beginner in AI, key skills to focus on developing: Mathematics and Statistics: A strong foundation in mathematics and statistics is crucial for understanding AI algorithms and models. Concepts such as linear algebra, calculus, probability, and statistics are fundamental to AI and machine learning. Machine Learning: Familiarity with machine learning concepts and algorithms is essential for AI development. Understanding supervised and unsupervised learning, regression, classification, clustering, and deep learning techniques is crucial for building AI models. Deep Learning: Deep learning has become increasingly important in AI, particularly for tasks like image recognition, natural language processing, and speech recognition.

Hay muchos lenguajes de programación en el mercado. Empezar con Python tiene los siguientes beneficios: - Una gran comunidad especializado en el campo de tu elección. - Lenguaje de alto nivel y muy fácil de aprender. - Hay una cantidad casi ilimitada de librerías y crear uno propio es también fácil. - Permite acceder y entender muy rápidamente la programación orientado a objetos. - Después de aprender Python el aprendizaje de otro lenguaje es rápido.

Diving into AI, a good first step is mastering a programming language, and Python is your best bet. It's straightforward, making it a good starting point, and it's widely used in AI for everything from machine learning to deep learning. As you grow, you might find the need to learn additional languages like Java for building software or R for data analysis, depending on your projects. The goal is to start with Python to build a solid foundation and then expand your skills based on your interests and the demands of your work. Regular practice will help you improve, so it’s important to keep coding and experimenting, for example by starting some personal projects.

I believe python, start there and then work your way up. Depending on what you want to achieve learn forms of sales, negotiations and communication. Even the greatest AI tool needs someone behind it.

La IA no deja de ser en parte las matemáticas y la estadística aplicada. Así que mejora tu entendimiento sobre los conceptos matemáticos y adquiere un libro de estadística. Recomendaría los 3 siguientes libros: 1- Statistics for Machine Learning Techniques for exploring supervised, unsupervised, and reinforcement learning models (Pratap Dangeti) 2- Practical Statistics for Data Scientists 50+ Essential Concepts Using R and Python (Peter Bruce, Andrew Bruce, Peter Gedeck). 3- Python for Probability, Statistics, and Machine Learning (José Unpingco).

For working with AI, deep theory isn't always needed, especially for practical tasks. But knowing the basics—like algebra, calculus, and statistics—helps you understand and apply AI tools better. This basic knowledge is enough to get started and tackle most projects. There are many online courses (on Coursera for example) that will introduce you to these topics. If you're curious or need to dig deeper into how AI works, then you can explore more complex concepts. Start with the essentials and learn more as you need.

Mathematics and Statistics: A strong foundation in mathematics and statistics is crucial for understanding AI algorithms and models. Concepts such as linear algebra, calculus, probability, and statistics are fundamental to AI and machine learning.

In my experience it depends which part of AI you intend on learning. As per usual there is no one silver bullet and it depends where you start at. I would strongly asses my mathematical understanding of linear algebra/statistical modelling/bayes and probabilistic and start filling in the gaps. If starting from the ground up and you just want to be able to comprehend the application of the background mathematical theory I would just focus on linear algebra specifically matrix multiplication and vector spaces. If you really want to understand optimization techniques such as backpropagation you need a fair bit of differential calculus. After going through some areas of AI each will have particular mathematics that require further study.

Entender las estructuras de datos es una parte vital para desarrollar aplicaciones en la IA. En general hay dos tipos de datos: - Los datos estructurados como los que se encuentran en fuentes de datos como SQL, json, csv, excel etc. - Los datos no estructurados como archivos de textos, videos, imágenes y audios etc. Además de los datos, que serien los ingredientes si hacemos similitud con el proceso de cocinar, necesitas los algoritmos. En general, los algoritmos son los pasos para hacer un buen plato, lo que seria una aplicación/software en el campo de la IA. Recomendaría el siguiente libro para entender los algoritmos más básicos en la IA: - Inteligencia artificial -Fundamentos, práctica y aplicaciones (Alberto García Serrano)

In AI, knowing your way around data structures and algorithms is key, but it's something you can pick up alongside learning Python. As you dive into coding, especially through personal projects, you'll naturally come across essentials like arrays, lists, and more complex structures like trees and graphs. Similarly, you'll encounter algorithms for sorting or searching as part of solving real problems. This hands-on approach lets you understand how to organize and process data without getting bogged down by theory. Focus on practical learning and application; the theory will make more sense as you apply it in actual projects.

Fundamental concepts in computer science and are essential for AI. Efficient Data Storage and Retrieval: Data structures allow you to organize and store data in a way that facilitates efficient retrieval and manipulation. Understanding different data structures, such as arrays, linked lists, trees, graphs, and hash tables, enables you to choose the most suitable structure for specific tasks. Optimized Algorithms: Algorithms are step-by-step procedures for solving problems and performing tasks. By studying algorithms, you learn how to develop efficient solutions to various computational problems, such as sorting, searching, graph traversal, and dynamic programming.

Specifically for AI you don't need to master all data structures but should be aware of their existence. To manipulate data you will work with varios forms of vectors/tensors in 1,2,3,4+ -d dimensions for this you should really focus on your arrays and lists. This should be your foundation for loading/manipulating data. Afterwards if you dive into decision trees you would obviously need to study the underlying data structure. Same story for Graph Neural Networks and Graphs, Greedy searches and greedy algorithms and so on.

After grasping core concepts like supervised/unsupervised learning and neural networks, hone in on an AI niche that fascinates you. Mastering relevant tools (like TensorFlow for deep learning or LangChain for LLMs) becomes essential as you delve into creating specialized AI applications. This targeted approach not only simplifies your learning journey but also empowers you to build practical, impactful AI solutions.

Deep Learning can be a sharp sword with two edges that helps you immensely but also handicaps your learning journey. You can solve most problems with sufficient data and a proper model that is why we have so much excitement for the domain. You can either start top down or down up according to your previous learning history. I recommend the latter where you write a model first with vanilla python and numpy and build upon it the layers/training strategies. There is also the other way around where you master a top level framework like keras/pytorch lightning and go down from there to tweek the smaller components.

El conocimiento del dominio y resolución de problemas son fundamentales para enfrentarse a cualquier tipo de aplicación en la IA. Generalmente, el primer punto de entrada es la colección de los datos. Luego, entender el contexto, los procesos de adquisición, y los conocimientos detrás de estos datos es imperativo para crear una aplicación fiable y útil. Hoy en día es relativamente fácil acceder a fuentes de datos fiables pero lo que llevar más tiempo es adquirir el conocimiento base de los datos con la exploración. Una herramienta que facilita este proceso seria la visualización de los datos además de las técnicas de estadísticas.

This is where your background and interests play a significant role. AI becomes especially powerful when it intersects with other areas, like healthcare, finance, or environmental science. Reflect on how AI can enhance your field or a domain you're passionate about. Leveraging your existing knowledge and focusing AI applications in areas you care about not only makes your work more impactful but also more fulfilling.

This can very immensely from person to person. If you have a scientist background or a domain related background start with that to properly analyse your data and maybe extract the relevant features/create proper validations for your model. If you don't have any domain knowledge be easy that it is not required. Try to adjust while you go your understanding of the problem and be careful when validating your results by yourself. Try to get a second domain experience that you are going in the right direction and have patience with yourself and learning that said domain.