About This Website
I have always been facinated with languge. I got my first exposure to language tools when I discovered WordNet in the late 1990s. It was amazing to see a formal database represnetation of words I used every day. Leveraging WordNet for writing and search was opened a new door for me. This helped me formalize my ability to visualize how to represent concepts in a graph.
I first started using generative AI seriously when OpenAI release their first tools in 2020. I had been using BERT to analize clinical documents, and now there was another model that would not only fill in a single word, but it would continually guess a set of next words.
Back then it was just GPT-2.
But it still had potential. I started blogging about
GPT-3 in September of 2020
as a way to Generate Detailed Lesson Plans for my STEM students.
Although my peer group of graph and NLP researchers were interested in using these models for task automation, we found that by adding detiled examples of what we wanted we could dramatically increase the quality of prompts. So I started teaching informal one-hour sessions on how to us GPT to perform various tasks that we had formally done by builing custom BERT models.
The more we worked with GPT-3 the more we realized the incredible number of tasks that this single model could do. The demmand for the class continued and I launched both a half-day and eventually a full-day of training for developers. These were often the most popular classes that were offerend by our training group.
Since those early days, the power of generative AI has continued to expand. But the role of writing good prompts is still critical for you to get the most out of your models.
I hope that you find this course useful and I would appreciate any feedback you have to help me make the course better.