Generative AI has become a prominent topic in today’s market buzz. There are thousands of books and materials available online that would easily help one to dive deep into this subject. However, understanding its evolution is challenging, just as tracing human evolution over millions of years. AI too transitioned from a simple academic discipline to today’s Generative AI over the last half-century.
Let’s delve into the major timelines and transition from AI to Generative AI:
1950Alan Turing introduced “intelligent machinery” and discussed deep learning, hindered by limited computing power and data availability.
1956The Dartmouth Summer Research Project on Artificial Intelligence convened, birthing AI as an academic discipline. John McCarthy coined the term “artificial intelligence.”
1959Arthur Samuel pioneered machine learning with a self-learning checkers program.
1960John McCarthy developed the LISP language for AI tasks. The following year among the first functioning examples of generative AI, the ELIZA chatbot was created by British scientist Joseph Weizenbaum. It was one of the first examples of Natural Language Processing (NLP) involved emulating the role of a psychotherapist, engaging in natural language conversations with humans.
1960to 1970s — Research on computer vision and pattern recognition advanced.
1980Recurrent neural networks (RNNs), were introduced in the field of natural language processing and used for language modeling tasks. RNNs can model relatively long dependencies and allow generating longer sentences.
1990to 2000s — Computing power and data collection and processing surged with the rise of internet.
2010Machine learning, neural networks, and deep learning gained traction given new opportunities to develop smarter and responsive systems.
2014 One of the fundamental breakthroughs in generative AI is the creation of Generative Adversarial Networks (GANs) by an American computer scientist Ian Goodfellow. It is an unsupervised machine learning algorithm that engages two neural networks that are in competition with each other. One network is a generative model that generates content, and the other is discriminative that tries to figure out whether it is an authentic sample or not.
2017 Another type of model that has played a significant role in the development of generative AI is the transformer architecture model. It’s a deep neural network algorithm.
The transformer architecture is applied in NLP, which has led to the creation of large language models such as BERT and GPT.
2018The first version of GPT (Generative Pre-trained Transformer) was created by OpenAI. It was a major breakthrough in generative AI and, trained on about 40 gigabytes of data and consisting of 117 million parameters, GPT paved the way for subsequent LLMs in content generation, chatbots and language translation.
2021DALL-E, an AI for generating and editing unique artworks and photorealistic images was launched.
2022 Stable Diffusion and Midjourney AI image-generating tools debuted.
2023 GPT-4 was released, consisting of 1 trillion parameters and is capable of receiving both text and image prompts.
2024OpenAI’s Sora a diffusion transformer was launched. It is a text to video model designed to generate videos based on descriptive prompts and can also extend existing videos forwards or backwards in time.
Generative AI technology has a brief yet significant history, commencing around the mid-20th century. However, its major breakthroughs occurred in the 2010s, propelling its rapid development to unprecedented levels.
The future of generative AI hold promises across various domains. As the technology advances, we can expect to see more sophisticated and versatile generative models capable of producing higher quality and more diverse outputs. This could lead to breakthroughs in creative content generation, personalized medicine, synthetic data generation for training AI systems, virtual reality, and more.
However, ethical considerations such as ensuring fairness, transparency, and responsible use will become increasingly important as generative AI continues to evolve.
