We Are In the Cambrian Explosion of Generative AI

Learn more from Ilana in Leading Your Department/Team in Understanding Generative AI and shape your school in adopting and adapting to new technologies.Offered August 5-9.

ilana saxe
Assistant Dean of Faculty
The Lawrenceville School

​This moment in generative AI feels a bit like the Cambrian explosion, a period of time about 530 million years ago, when there was a sudden burst of diverse life forms. Need an extra set of appendages or ten? Look no further than the Cambrian explosion (or Dall-E). To be fair, the term “explosion” is a little misleading, since it can conjure the image of a magician snapping her fingers, making species appear out of thin air. In reality, the “Cambrian was preceded by many millions of years of evolution” (Berekley, n.d.). Similarly, Generative AI has actually been in development for years (Hawley, 2023), but now it seems like it is popping up everywhere and in so many new magical forms! ​

The thing is, Generative AI isn’t magic. It’s people. Behind every large language model are millions of pages of information, which contain the author’s biases as well as the biases of those who chose what to include and, just as important, what to exclude from the data set. So what are we as academic leaders to do? Do what we do best: be a model. Model for our students and colleagues how to use AI transparently, responsibly and collaboratively.

Transparency: 
Ethan Mollick writes about cyborgs, people who use AI for work, yet refrain from discussing their use of this tool with colleagues. In a 2023 article, Mollick writes, “there are at least three reasons these cyborgs stay secret. But they all boil down to the same thing: people don’t want to get in trouble.” However, when you have a school culture that celebrates new ideas and experimentation, you can have an open discussion about when and where to incorporate AI into your work and when not to. Similarly, teachers who use AI to lesson plan and refuse to pull back the curtain on this process miss out on an opportunity to teach students how to use AI. I recently asked Gemini to “please pretend to be a lipid soluble hormone, specifically testosterone, and write a song in the style of Taylor Swift,  from the perspective of the hormone, describing what happens once you are released (and from where), how you enter the cell, moving through the cell membrane (because you are lipid soluble) and then bind to a receptor protein to affect gene expression.” The song it created was hilarious and memorable, but was it accurate? As I was initially drafting the instructions for the class, I told them not to worry, I had checked the lyrics and they were biologically valid. But then I paused. Had I kept this in the instructions I would have missed an important teachable moment: pressing students to check for accuracy. Instead, I told my students I used AI to write several songs, but we couldn’t trust that what was created was accurate. Checking for accuracy became their job and they jumped right in.  By modeling how I used AI to create the lesson along with sharing possible pitfalls of the tool, I was not only helping students understand signal transduction pathways, I was teaching a transferable skill: approaching AI-generated content with a discerning eye. 

Responsible Use:
The Association of Academic Leaders February AI meetup, hosted by Sarah Hanawald, highlighted advice from Venable, the law firm, on things to consider when crafting your school’s policy when it comes to generative AI (Pass & Sykes, 2024). There are a myriad of issues to address: from student privacy to when and where AI is used. Should there be different standards for use on a document that is generated right before class, solely for a teacher’s use during that period, compared with an article shared on the school website? I would think so, but I’m also not a lawyer. Understanding how a generative AI tool stores and uses the data that users feed into it is important, especially when it comes to concerns about copyright and student privacy. At the Loomis Chaffee AI Conference: Navigating Uncertainty, Embracing Possibility: AI in Independent Schools, one of the presenters shared how he protects student privacy by replacing the names of his students with famous musicians or athletes on work that he uploads and ask AI to evaluate. Sharing hacks like this can be a stop gap for protecting student privacy, but schools need to carve out time to work on these policies as well as educate faculty and staff on how to engage with these policies. Given the pace of change in what is possible with Generative AI, these policies need to be living documents, regularly reviewed for compliance with local laws.

Collaboration:
Given the explosion of generative AI tools, it can be easy to feel overwhelmed and like it is impossible to keep up with what is happening in the field. That tsunami of possibilities can drown curiosity courtesy of choice paralysis. How do you even decide which AI tool to use or to team up with? Take advantage of the saying ‘many hands make light work’ by having an AI Discovery committee made up of constituents that represent all aspects of your school community: faculty, students, staff and even current parents and alumni. Ask each member of the committee to focus on one tool or follow one critical thinker in the field and every other week update a launch pad with a “tiny gem” or notable update on this tool. Prior to meeting as a group, members can read over the launch pad to see what fellow members have added. Now this is where you might ask “couldn’t AI just do that for me?” Yup, it sure could. But, I would argue that in this case having the person who vet those sources be someone who understands your school culture and mission matters. 

The current boom in generative AI presents both exciting opportunities and significant challenges for the academic community. As educators, we have a responsibility to embrace transparency, promote responsible use, and foster collaboration within our schools to harness the power of AI while mitigating potential risks (a task that makes lunch duty seem like a breeze!). By modeling critical thinking and open communication, we can empower future generations to navigate this rapidly evolving technological landscape and ensure that AI serves as a tool for learning and progress,  expanding the realm of what is possible in education and promoting critical thinking.

Sources:
Hawley, M. (2023). The Complete Generative AI Timeline: History, Present and Future Outlook. CMSWire.Com. Retrieved February 29, 2024, from https://www.cmswire.com/digital-experience/generative-ai-timeline-9-decades-of-notable-milestones/

Caryn G. Pass & Ashley E. Sykes. (2024, February 29). Emerging Employee and Student Handbook Updates for the 2024-2025 School Year | Insights | Venable LLP. Venable.Com. https://www.venable.com/insights/publications/2024/02/emerging-employee-and-student-handbook-up

UC Museum of Paleontology Understanding Evolution. (n.d.). The Cambrian explosion. Retrieved February 29, 2024, from https://evolution.berkeley.edu/the-cambrian-explosion/