Introduction

By necessity, this book is widely interdisciplinary, bringing together insights from scholarly work understanding "understanding," social action, social systems, the social psychology of belief, the philosophy of science, the sociology of belief systems, research ethics, ethics of privacy, and of interaction, clinical psychology, the technical intricacies of LLMs, frameworks of knowledge management, automated proof-checking, to name some of the most important fields of knowledge involved.

As you can see, this textbook cannot be built by just one person. I (Alec McGail) am writing this now to start the endeavor in a free and transparent manner, very much in the spirit of what's discussed in Section 2: What's Ethical?. So, anyone who feels they have something they can contribute to the endeavor should reach out to me (am2873@cornell.edu), or just go ahead and make changes.

If you'd like to follow my process in working through this class, follow my Twitch channel and YouTube channel.

Here, you will embark on an intellectual adventure, blending the theoretical intricacies of intersubjective thought with hands-on training in Large Language Models (LLMs). By the end, you won’t just understand the mechanics of these digital marvels; you will be the craftsman behind their creation.

For the intrepid scholar and the visionary educator alike, this journey promises a harmonious blend of robust theoretical foundations and cutting-edge practical applications. Each week unfurls a new layer of understanding, from ethical considerations to technical mastery, all culminating in a capstone project where you breathe life into your very own chatbot. This isn't just a course; it's a call to be at the forefront of a sociotechnological revolution, with the power to shape discourse, challenge beliefs, and unite our ever-evolving global community. Get ready to be both the student and the pioneer, charting the path for the next wave of societal evolution.

Disclaimer: "Chatbots for Social Change" has been collaboratively developed with the aid of ChatGPT, a product of OpenAI's cutting-edge Large Language Model (LLM) technology. The utilization of ChatGPT in the creation of this WikiBook is a practical demonstration of the subject matter at the heart of this course. As learners delve into the complexities of collective cognition, LLM training, knowledge management, and social interaction, they are interacting with content that has itself been influenced by the advanced technologies under discussion.

This recursive element of the course design illustrates the dynamic and evolving interaction between human intellect and artificial intelligence. It's an embodiment of the dualities and partnerships that can emerge when the creative capacities of humans are augmented by the meticulousness of machine intelligence. This partnership is indicative of the immense possibilities and responsibilities that come with the integration of such technologies into the fabric of our digital era. Understanding, leveraging, and steering these advancements remain a central theme and imperative throughout this WikiBook.

Independent Learning

Independent consumption by definition means you can do whatever you want with your time and this book. So do that! Read as much or as little as you like, skip around, and don't be shy about asking questions.

If you are serious about learning the content, you will have to devote significant time. I recommend setting aside consistent time every week, and work through the book section by section. Do the prototypes yourself, and contribute to the wiki book. And email me (am2873@cornell.edu)!

Teaching the Course

I am developing this textbook in approximately the same structure I imagine one would teach a 9-week intensive (perhaps summer-) course.

Weeks 1-3: Sections 1-3, which are largely theoretical, could be presented in the first three weeks. This makes it a whirlwind tour, but the textbook allows students to dig deeper at their discretion. At the end of the second section What is Ethical?, I imagine students would draft an IRB proposal for a intervention they would like to conduct. This serves to focus students on what they'd like to do with the technology before the third theoretical week How Do We Do It? and the following technical weeks which prepare the student for their own prototyping.

Weeks 4-5: The next two weeks can be spent on the technical details of LLMs and various other relevant technologies. Students can choose a topic from the textbook to explain to the class, or choose to research a new one and write a wiki chapter.

Weeks 6-8: The next three weeks would involve hands-on prototyping based on the subject matter. This gives a nice fail-fast mentality, and avoids "scope creep," which can easily result in never getting off the ground. This would be wildly benefited by a user-friendly package which includes high-level access to capabilities for everything mentioned in this book.

Week 9: The final week can be used to reflect on the course material, and allow students to present what they were able to do, what challenges they faced, and their ideas for further development and use of these technologies. If they feel they can contribute to the code-base, this would be a good time to submit their pull-requests.