I would have never thought that wrapping your head around the idea behind AI may be so challenging, awe-inspiring, and eye-opening! At some point, I understood that it helped me understand myself better.
I called my Blog AI Odyssey because the process of unraveling AI feels like a huge journey. Furthermore, there is so much information, that sometimes I have the impression that I am just starting my quest. In other words, I have just cracked open the door to peek out onto the journey awaiting me! It is just the beginning. Below I have included a couple of insights that helped me get started!
What is AI?
AI is a branch of computer science that deals with the creation of intelligent agents, which are systems that can process vast amounts of data to identify patterns, make informed decisions, and generate new content. For me, AI seems like a vast library of knowledge, where machines access and assimilate a mind-blowing amount of information to create something new.
The work of AI is rooted in the ideas of Machine Learning and Deep Learning.
"Machine learning and deep learning are methods where computers learn to do tasks without being programmed explicitly. Machine learning uses algorithms to learn patterns and make decisions, while deep learning uses neural networks to imitate the human brain and solve complex problems." ChatGPT
I found the following illustrations from a Google Course "Machine Learning and Artificial Intelligence" very useful. Slide to see all three images!
What is Generative AI?
What makes AI unique is that it can learn autonomously, which is inspired by the human brain. This is where approach the term Generative Artificial Intelligence. It is a type of Artificial Intelligence that creates new content based on what it has learned from existing content. The process of learning from existing content is called training and results in the creation of a statistical model. There are different kinds of Generators. I have united some of them in the illustration below:
When given a prompt, it uses this statistical model to predict what an expected response might be–and this generates new content. It learns the underlying structure of the data and can then generate new samples that are similar to the data it was trained on. As previously, a generative language model can take what it has learned from the examples it has been shown and create something entirely new based on that information. Hence the word “generative!”
Are machines like humans or are humans like machines?
The idea that AI generates new content by making predictions after analyzing a vast amount of data fascinates me the most! Some psychologists and neuroscientists believe that this is exactly how our human brains work. Hannah Critchlow in her book “The Science of Fate” suggests that our choices, no matter how big or small, are the results of a long and elaborate series of predictable decisions that take place in our brains on a subconscious level.
We might erroneously think that we decided to have eggs and bacon for breakfast “out of the blue”, but, in reality, our choice could have been predicted. If we could consciously analyze all the factors, like food that we have in our fridge, our mood of the day, our heart rate and volume of stomach fluid, and what we had had for dinner the day before, we would be able to predict our breakfast choice with striking accuracy! The same rule applies to career choices, selection of partners, or picking favorite vacation spots! This is all statistical data! And we are more predictable than we think.
So, how can ChatGPT write stories with unexpected endings?
I once asked ChatGPT to create a story about an 8-year-old boy, a dog, and a mosquito. Surprisingly, it generated an unexpectedly kind fairy tale where a mosquito plays with an 8-year-old boy by tickling rather than biting his hand. You may argue that ChatGPT came up with a plot that was not predictable, right? Because the most predictable scenario would be a mosquito biting a boy. But I think that ChatGPT had been programmed in such a way that it knows: if this is a story, it should have an unexpected scenario to make it more fun. Thus, it selected the most unpredictable of all choices available. Why did ChatGPT make a kind story? Because he might have been programmed to create “kind” stories for kids aged 8!
Can AI help us get better?
So, with the outbreak of machine learning, we as humans get a chance to get a better understanding of ourselves! If our choices are as predictable as those of an AI machine, maybe we can become even more mindful. If we become more conscious of our choices, and make an intention to be kinder to ourselves and others, we will not fear Artificial Intelligence, which operates on algorithms designed by Humans!
