# Artificial Intelligence vs Machine Learning vs Deep Learning: What’s the difference between them? > AI is the broad field, ML a subset, and Deep Learning uses neural networks. How to tell them apart and where each one applies. Source: https://www.erathos.com/en/blog/artificial-intelligence-vs-machine-learning-vs-deep-learning Em português: https://www.erathos.com/blog/artificial-intelligence-vs-machine-learning-vs-deep-learning Published: 2022-10-11 Category: Data Science & AI ![Venn diagram showing the relationship between Artificial Intelligence, Machine Learning, and Deep Learning](https://cms-media.erathos.com/Mj84pNfgEzph8GY20vKUXVHD7gA-1.png) The concepts of Artificial Intelligence, Machine Learning, and Deep Learning are often used interchangeably, but they have key fundamental differences. In this article, we'll dive deep into what those differences are and the main applications of these technologies, especially for data science. ## What is Artificial Intelligence? Can machines think? This question, originally documented in an [article by Alan Turing](https://academic.oup.com/mind/article/LIX/236/433/986238) back in 1950, was one of the catalysts that helped drive major breakthroughs in computer science. This is especially true when it comes to Artificial Intelligence (AI), which aims to apply the rational elements of the human mind to robots and other computational systems. The history of AI development has been marked by a series of hurdles, breakthroughs, delays, and waves of high investment and funding winters. However, the last two decades have seen remarkable progress, powered by advancements in robotics, the internet, digital media, and more high-level, less complex programming languages. This field is still rapidly evolving, with virtually infinite possibilities for application. What used to be extremely difficult to understand or see in real-world scenarios is now part of everyday life. For example: software that generates [images and digital art](https://openai.com/dall-e-2/) or even [videos](https://makeavideo.studio/) from natural language prompts; highly intelligent [digital assistants](https://tecnoblog.net/responde/o-que-e-a-alexa-ou-melhor-quem-e/#:~:text=Como%20funciona%20a%20Alexa%3F,usu%C3%A1rio%20vai%20executar%20um%20comando.); [translation software](https://www.deepl.com/translator) capable of interpreting figurative context and adapting it across languages; among other simple and complex applications. How AIs work depends on how they are programmed, but they can be broadly divided into three main categories: **Supervised**: In this type of AI, the output is directly defined by the developer. For example, an AI trained to recognize and flag specific items, objects, people, or animals will eventually be able to do so autonomously, based on the parameters it originally learned. **Semi-supervised**: Here, the model aggregates the information it has learned to output the expected result, aligning with what the person managing its learning process expects. **Unsupervised:** In this category, AIs can independently draw parallels between different items, objects, or samples without the person training them specifying a desired output. Within the broader scope of AI, we find Machine Learning and Deep Learning. In other words, both are subfields of AI, even though they represent distinct concepts. ## What is Machine Learning? Machine Learning is a field dedicated to studying and developing algorithm-based techniques and methodologies that allow machines to improve their ability to perform tasks autonomously. Broadly speaking, they can be applied across various domains, ranging from speech recognition (like Amazon's Echo Dot, the famous Alexa) to building highly accurate statistical models and predicting scenarios through mathematical optimization and data mining. These applications can be leveraged to improve data performance and drive optimizations across all business units, making it an incredibly relevant field of study for entrepreneurs and tech leaders. ## What is Deep Learning? Within Machine Learning, we have Deep Learning, the process of teaching computers to think by association and learn by example, mimicking how humans naturally acquire knowledge. Through Deep Learning, computers learn to recognize and classify tasks by identifying patterns in voice, images, video, and text. Some everyday examples of Deep Learning include self-driving cars, mobile digital assistants, aerospace defense systems, object identification in satellite imagery, and even medical research for genetic markers and cancer detection. ![nW3E4Xvg9tnN7pvfcMWbbdJT0Q4.jpg](https://cms-media.erathos.com/nW3E4Xvg9tnN7pvfcMWbbdJT0Q4-1.jpg) ## Wrapping up... The rapid advancement of Artificial Intelligence is a field that sparks curiosity and interest in everyone. Over the past few decades, this technology has evolved at an exponential rate, which also led to the rise of specialized subfields within AI, namely Machine Learning and, subsequently, Deep Learning. While many people confuse these terms, they are distinct areas of research within the broader scope of AI. Their applications are having an increasingly profound impact on our daily lives, powering smarter mobile devices, safer vehicles, household appliances that simplify routines, and medical breakthroughs that improve our quality of life. Each of these fields holds infinite potential for expansion, especially for businesses looking to build a more data-driven culture. To learn more about applying these technologies to your business, visit the Erathos blog, where we regularly publish content on how to build a data-driven organization.