The new era of generative AI

In recent years, one of the most frequently used and widely discussed terms in the world of technology is artificial intelligence (AI). Almost every day we come across news about this announced revolution; however, it is rarely specified which type of artificial intelligence is being referred to, since there are different kinds. What is generating so much excitement is generative AI. Let’s therefore try to better understand what generative AI is.

What is generative AI?

Generative Artificial Intelligence, or Generative AI, is a technology that became widely known thanks to OpenAI’s ChatGPT software, a startup owned by Microsoft. Today, there are many other software tools of this kind. Generative AI can improve the performance of various activities for both individuals and companies, such as the production of texts, images, and standard software code. Its use speeds up work and, to some extent, can be described as creative, thanks to the combination of large amounts of sources and data used.

According to McKinsey’s definition, Generative Artificial Intelligence describes algorithms that can be used to create new content, including audio, code, images, text, simulations, and video.

Generative AI systems fall under the broad category of General Artificial Intelligence (AGI) and machine learning (Machine Learning, ML). They have the potential to change the way we approach content creation in applications such as design, entertainment, e-commerce, marketing, scientific research, and human resources.

It is clear that we are in a moment of major evolution, but we are still at the beginning, and we will need to wait a bit longer to better understand what the real developments of generative AI will be and what its consequences, opportunities, and risks will be for our lives and jobs. What is clear is that training in this field—for example with a Master in Artificial Intelligence & Machine Learning for Business—is a choice that will open up interesting career prospects.

How does generative AI work?

Generative AI software starts from requests or descriptions (prompts) formulated in natural language by the user (human or software) and, as a result, generates text from text (Text-to-Text), images from text (Text-to-Image), or even images from images (Image-to-Image). The results of these systems are combinations of the data used to train the algorithms.

Due to the enormous amount of data used to “feed” the software (the GPT-3 system on which ChatGPT is based was trained with 45 terabytes of text data), the results may appear “creative.” In reality, what they generate is a combination of sources, but given the vast amount of processed data, the output can feel truly new. After all, reprocessing can also be considered a form of creativity.

Understanding how generative AI works is not easy for non-experts: the idea is that, through feedback and training, these intelligences continuously improve. For example, ChatGPT technology could be described as an example of a Generative Adversarial Network (GAN). However, this is debated, because some experts argue that ChatGPT is a Transformer (GPT stands for Generative Pretrained Transformer) rather than a GAN. What does this mean?

A Transformer is a deep learning model used in the field of NLP (Natural Language Processing), where outputs are generated from a reprocessing of previously stored information. GANs, on the other hand, are a type of artificial intelligence algorithm that uses two neural networks competing against each other to generate images, sounds, text, and other types of data. The first network, called the “generator,” tries to create fake images or data that look real; the second, called the “discriminator,” tries to identify whether the images or data are real or fake.

The two networks compete with each other: the generator tries to produce increasingly realistic data, while the discriminator improves its ability to detect whether the data is real or fake. Over time, the generator becomes better at producing realistic data that fools the discriminator, while the discriminator becomes better at detecting fake data. The goal of a GAN model is to optimize deep learning and avoid shallow generalization errors due to data scarcity.

Examples of generative artificial intelligence

There are several very interesting examples of generative AI:

• ChatGPT is software that simulates and processes human conversations; it can generate text that responds coherently to user questions.
• Midjourney, Stable Diffusion, and DALL·E create images from text.
• Make-A-Video converts text prompts into short videos.
• Synthesia allows the creation of highly realistic “AI avatar” videos capable of speaking 120 languages.
• MusicLM generates music from textual descriptions.

Uses of generative artificial intelligence in companies

IT companies can leverage generative AI to instantly generate code. Organizations that need short marketing texts or technical manuals also benefit from it. However, currently generative AI is most effective in producing standard content (such as emails, CVs, or manuals).

Generative AI offers design companies a faster and more efficient way to create and modify projects. Algorithms can be trained with large datasets, such as images of existing products, and then generate new designs and models that meet defined criteria, as well as modify and personalize existing designs, creating new variations and options. Applications range from the fashion industry to automotive, as well as building design and other architectural works.

In the retail sector, generative AI is used for product and content personalization: emails or product recommendations, promotional content (ads and posts), website and mobile app design. AI can also generate descriptive text for each product in an e-commerce catalog. Changing the visual characteristics of products or describing them in videos is another application area.

Companies that want to use generative AI can either use the technology as it is or train it by adding their own data and models. However, this requires large amounts of data and significant capital. Even so, SMEs can also use generative AI to plan production and distribution and improve customer experience through personalized content production.