From The Spinning Jenny to ChatGPT
Putting the rise of modern AI into a historical perspective
--
I have had a passion for artificial intelligence since the 1980s. It was one of the reasons why I studied machine learning and robotics over 20 years ago. My final project for undergrad, for instance, was a system for analyzing whether fish had parasites using numerous sensors connected to a neural network. But I am not writing this story to go down memory lane but to reflect upon what the rise of modern AI means to human society today and in the near future.
Over the last decade, there has been a string of breakthroughs. We have gotten speech to text dictation, which finally works well. Image recognition has finally become useful in everyday life. Many of you have probably used face recognition in photo album software to organize photos. Semi autonomous driving has made impressive advances and gets used by countless people in their daily life.
AI Art Generation
The latest breakthroughs have been in using AI such as Stable Diffusion and DALL-E 2 to create art. These AI image generates can be used in a variety of ways. You can create an image based on a written prompt, or you can take existing images and make variations or adjustments. Below is an example of DALL-E 2 creating a variation on the famous Vermeer picture, Girl with a Pearl Earring.
The two images below are both generated from the same textual prompt using Stable Diffusion:
A large cabin on top of a sunny mountain in the style of Dreamworks, artstation.
AI Text Generation
Even more stories have been written about ChatGPT which is a sort of advanced AI chatbot launched by OpenAI in November 2022. You can interact with ChatGPT and unlike the typical chatbot it remembers which you have told it making it more like a real conversation. ChatGPT is trained on a huge amount of…
