While AI is undeniably evolving more and more every day, the debate around AI’s possible consciousness is still very much active. Indeed, the cutting-edge AI tech GPT-3 (Generative Pre-trained Transformer 3) is stepping closer to being described as conscious.
GPT-3 is an autoregressive language model, using deep learning to produce human-like text and is apparently able to create content better than anything else ever made. The AI can answer questions, write essays, translate various languages and create computer code, all of this without any human input, or almost. Hence, GPT-3 is, for now, the biggest artificial neural network ever created.
However, there have been some signs allowing researchers to claim that AI is developing a consciousness. This is partially due to the fact that GPT-3 has been trained on 45 terabytes of text data and can thus process around 45 billion times the number of words a human being absorbs over a lifetime.
Whether GPT-3 is actually conscious or not is up to everyone to believe, yet there is the question of how far will AI go? This then brings the topic of Artificial General Intelligence (AGI), where AI evolves from a linear ‘existence’ to one where it can reason in the same way that humans can. Although we are not there yet, San Francisco researchers believe that GPT-3 opened the way to this new future.
The ever-evolving world of AI is on its way to become ever more powerful and interesting. Thus, AI might be, in time, the key to take on the most complex challenges we face. However, it can also go the other way around and be used for dangerous and harmful activities such as surveillance, hacking, and warfare.