"AI made simple: a clear perspective on types and real-world applications"

Introduction Artificial Intelligence (AI) has been evolving rapidly, especially since the launch of ChatGPT, which has had a profound impact worldwide.

Released in November 2022, ChatGPT reached 100 million monthly users within two months and now (as of November 2024) boasts 200 million weekly users, or 800 million active monthly users, making it the fastest-growing application in history. [1]

With so much hype and conflicting information, it is essential to understand the basics of AI before delving into its innovations. In this article, we will distinguish traditional AI from generative AI, clarify its relationship with machine learning (ML), and provide real-world examples of each.

1. What is Artificial Intelligence? Artificial Intelligence (AI) refers to the capability of machines and computational systems to perform tasks that traditionally require human intelligence, such as reasoning, learning, perception, and natural language interaction. These technologies enable computers and devices to simulate problem-solving skills and human-like intelligence.

It is important to note that not all technologies that appear intelligent are genuinely AI. Systems that merely follow predefined instructions, such as algorithms that calculate the sum of two numbers, are not considered AI. True AI involves the ability to learn and adapt to new information, going beyond fixed instructions.

Today, AI is embedded in nearly every aspect of our daily lives—from personalised recommendations to medical diagnoses. What's most fascinating is its diversification into various specialisations, making it an incredibly flexible and adaptable tool—a true digital Swiss Army knife.

2. Traditional Artificial Intelligence vs. Generative Artificial Intelligence

2.1 Traditional or Classical Artificial Intelligence Traditional AI involves algorithms that execute specific tasks, such as pattern recognition or rule-based decision-making. These systems are effective in well-defined problems but do not generate new information or content.

This type of AI has been with us for quite some time, often unnoticed. It is embedded in software we use for work, apps on our smartphones, and many other facets of daily life.

Occasionally, I hear claims like, “My company uses AI for...” in conversations or presentations. I am convinced that what they often mean is traditional AI—which is no longer a differentiator—rather than generative AI, as they might imply.

Examples:

• Fraud detection systems used by banks and credit card issuers, which analyse transactions to identify suspicious behaviour. [2]
• Recommendation systems on platforms like Netflix, Spotify, or YouTube. These systems analyse user behaviour to suggest films, series, or music but do not create new content; they simply organise and recommend what already exists in the catalogue. [3,4]
• Navigation systems such as Google Maps or Waze, which calculate optimised routes.

2.2 Generative Artificial Intelligence Generative AI, in contrast, is a subfield of AI that uses advanced models to create new content, such as text, images, or music. These models learn from vast amounts of data and can produce original outputs.

Unlike traditional AI, which is confined to specific tasks, generative AI uses neural networks to "learn" patterns and generate more flexible and creative results.

Examples:

• ChatGPT (https://openai.com/chatgpt), which generates textual responses based on extensive data training.
• Google’s Bard (https://bard.google.com/), which aims to answer complex questions with real-time generated responses.
• Perplexity AI (https://www.perplexity.ai/), which provides detailed and well-researched answers to queries.
• GitHub Copilot (https://copilot.github.com/), which helps programmers write code faster by suggesting code snippets based on previous patterns.

The Role of Algorithms and Machine Learning (ML) AI is fundamentally built on algorithms—sets of rules that guide decision-making. However, not all algorithms qualify as AI; many follow fixed instructions without the capacity to learn.

Machine Learning (ML), a subfield of AI, introduces a learning approach based on data. Instead of being explicitly programmed for every task, ML systems "learn" from data, adapting and making decisions based on identified patterns.

Examples:

• Amazon's recommendation system (https://www.amazon.com/), which analyses browsing and purchase data to suggest products.
• Spotify (https://www.spotify.com/), which analyses listening history and musical preferences to offer personalised recommendations.
• OpenAI’s DALL-E, which uses ML to generate images from textual descriptions, exemplifying the intersection of both technologies.

Final Thoughts In this article, we explored the foundations of Artificial Intelligence, from basic concepts to its latest innovations. Reflecting on this, how do you assess AI's impact on your sector or professional life? What challenges and opportunities do you foresee as the technology advances?

Share your thoughts in the comments, and let’s continue this conversation.

This is just the beginning of our journey. In the next article, we’ll delve into more advanced aspects of AI, including how neural networks function and the impact of generative AI on areas like art and content creation. I aim to conclude this mini-series with a third article on AI applications in Logistics and Supply Chain.

Stay tuned, as we dive deeper into how these technologies are shaping the future of creativity and innovation.

--

Sources

[1] https://www.demandsage.com/chatgpt-statistics/

[2] https://www.ibm.com/br-pt/trusteer

[3] https://medium.com/design-bootcamp/how-spotifys-ai-driven-music-recommendations-revolutionize-user-experience-bedbcf4f898a

[4] https://medium.com/@zhonghong9998/personalized-recommendations-how-netflix-and-amazon-use-deep-learning-to-enhance-user-experience-e7bd6fcd18ff