I got a notification in LinkedIn for a position titled, Senior Writer Editor. It was with a company called Outlier and the goal seemed to be to improve AI prompt responses. It sounded interesting and aligned to the type of work that I was interested in pursuing. After submitting my resume, I was given the go ahead to start the onboarding training to learn more about the position. The training started off interestingly enough. I was given this doc to explain how I would be rating prompts. Basically, I was given two responses to prompts and had to determine which was better and give a rationale based on criteria provided. This part of the training was manageable and enjoyable. I liked picking apart prompt responses for elements of quality. The training identified responses along a spectrum of creative to logical/factual. They also provided rubrics that were helpful for each type of writing. Here are some examples. Once I completed a few of the rating exercises, I was then given the instructions that I would be challenged to write a response to a prompt based on my expertise in K-12 education to help train the models. Here is the prompt that I received:
Delving into Gardner’s Multiple Intelligences and recent findings in neurodiversity, how can educators ensure they are differentiating not only based on ability but also on cognitive processing differences? Discuss the interplay of content, process, product, and learning environment in this context. When I showed it to my husband, he said it sounded like something you could write a PhD thesis on and I couldn't disagree. But I decided to go ahead and give it a shot thinking I would only have to write the one. I wrote what I felt was a decent response and it took me a couple of hours. Click here to read my response. When I finished, I was provided another prompt to respond to and that was where I stopped. Who knows how many more prompts I would have had to respond to in order to finish the training? It was clear at that point that I would be required to do a lot more work than it was worth. Nonetheless, this is a new movement in AI, getting the expertise of real humans to make the AI models better. It was an interesting exercise in being an AI Chatbot and I don't regret doing it. However, I will be wary of future opportunities such as this and you should too!
Educational, school and district leaders are scrambling to come up with policies and guidance regarding AI in schools. Students especially high school and college students are already using AI and there is no detection system available to adequately assess when AI has been used. Luckily there are a number of organizations working together to provide leadership and guidance. TeachAI has created an AI Guidance for Schools Toolkit that is the result of 60+ global organizations working together. It is designed to "help local, state, and national education systems worldwide develop guidance on the responsible use of AI, ensure compliance with relevant policies, and build the capacity of all stakeholders to understand AI and use AI effectively. " It provides a framework for implementing AI in an educational system and many resources for policymakers and educators to consider.
Code.org, ISTE, Khan Academy, and ETS have partnered together to create an AI 101 page to help educators think about how to use AI. There are a series of 30 minute videos that provide wonderful guidance in terms of thinking about using AI for teaching and learning and specific examples of teachers using generative AI to create content for their classrooms. This intro video below of Hadi Partovi of code.org and Sal Khan of Khan academy is a great place to start to think about all of the risks and benefits of AI in schools
In addition to the series of videos for educators, there are high quality videos explaining various topics related to AI and a growing collection of professionally designed curricula that students and teachers can access from Code.org. There is a link to ISTE's course for educators on generative AI and two AI tools specifically designed for students: ETS has developed a writing tutor for students and Khanmigo is Khan Academy's chatbot designed specifically for students.
The AI toolkit and AI 101 page offer thorough and well thought out ideas for implementing AI into school systems and both leave space for room to grow as the field of generative AI advances. Prompt Engineering has come onto the scene as an important means to use generative AI to its fullest. What and how you ask for information in a generative AI app can play a big role in the information you get. One big suggestion by many is to just play around with it and try it out. You can go to ChatGPT and just type in a prompt or click on one of the prompts that are listed. Because generative AI uses natural language models, there is no learning curve for initial exploration. It's also a good idea to compare different generative AI apps such as BIng Chat which uses the premium version of ChatGPT and can deal and has the most up-to-date information available or Perplexity.ai that includes sources with the information provided. Another possibility is Google's AI assistant known as Bard. The screenshot below shows how Perplexity.ai provides sources. The next step would be to get the AI assistant to refine the information it provided. For instance, in the above example about dolphins you might want to know more about the ways dolphins show self-awareness. You can continue to refine content and chat assistants do remarkably well with follow-up prompts remembering the history of your interactions.
Chat assistants have proven to be quite remarkable in the education space for lesson planning, assessment construction and differentiation, and there are a number of sites that offer pre-written prompts for educators. Code.org has two great prompt libraries the first one titled LLM prompts for educators. It offers a collection of prompts organized by beginner, intermediate and advanced and provides useful guidelines for creating prompts. The second library include prompts for using with students called AI prompts for transforming student learning. Another great place for educators to see a wide variety of prompts is AI for Education's prompt library. On this site prompts are organized by type such as lesson planning, administration and professional development. There are also a growing number of courses available to help one learn how to write prompts. I recommend the free course, Innovative Teaching with ChatGPT, to get started. Vanderbilt professor Jules White uses very teaching specific prompts to show how to create and refine lesson plans and activities as well as differentiate for different types of students. I recently took another course by White called Prompt Engineering for ChatGPT which takes a deeper dive by exploring some of the patterns that are useful in creating different types of prompts. AI is absolutely going to change the educational landscape and an easy way for educators to get started is to try their hand at prompt engineering. I think they will immediately find that chat assistants have the potential to really save them time. image source: https://medium.com/the-ai-education-project/introducing-the-ai-education-project-3c1f1fc31fd2 As I continue my journey exploring AI and its implications for teaching and learning, I spent some time reviewing the curriculum available at The AI Education Project. Their site has free curriculum available for students and educators as well as information for advocates. They have partnered with some of the big tech companies including Google, OpenAI, Microsoft and GitHub and have a mission to create equitable learning experiences that teach foundational AI skills. There are high interest, flexible lessons and activities that range from 5 minute warm ups to a semester long Introductory course. The AI Education Project implements culturally relevant pedagogy and project-based learning as a foundation for their curriculum and the content choices reflect a broad range of topics that teachers of any subject matter can find relevant.
AI Snapshots offer 180 five minute warmups organized by the four core subject areas: English, Math, Science and Social Studies. Each warm-up starts with a slide that asks students a thought-provoking question or design challenge. Then there is a second slide titled: Things You May Have Considered. That helps students and teachers explore the topic more deeply. It's a great way to get students to begin to think about the complexities and impact of AI in various disciplines and aspects of our lives. There are also AI Challenges that students can work through on their own that challenge students in timely tasks such as proving they are smarter than ChatGPT and improving their TikTok algorithm. These are wonderfully engaging independent lessons for curious high schoolers to try. For Computer Science and Technology teachers who are interested in bringing AI into their curriculum, the AI Education Projects offers a Project Dashboard that provides project-based learning on a variety of topics related to AI. One of my favorite projects on the dashboard is The 29 A.Is of Washington D.C. where students follow the journey of individual citizens and see how their lives are impacted by AI. It is a memorable, equity-focused lesson that drives home the problem of bias inherent in AI systems. The Intro to AI course is an incredibly thoughtful and well-designed course that provides foundational skills in AI while having students create their own AI recommendation system using Hugging Face. The course includes lesson plans, a teacher's guide, a slide deck and a student workbook. While this course is recommended for 10 weeks, it could easily be built out to last an entire semester. This course is one of the best examples of culturally relevant pedagogy in the field of computer science that I have seen. It gets students to consider AI in ways that are based in the real world. It has them explore biases inherent in data and gives students ample choice to explore their own interests. Furthermore it provides teachers with explicit guidelines to teach the course in a way that makes it accessible to those who may feel a bit intimidated to teach a course in AI. Finally, the AI Education Project offers live professional development and toolkits for educators and advocates who are interested in getting AI implemented in their classrooms, schools, and districts. The AI Education Project is doing incredible work in the field of equity focused and civic-minded computer science education. I highly recommend it as a place to go to find curriculum and guidelines related to teaching AI. Image Source: https://cs.illinois.edu/broadening-participation-computing/programs/ai4all AI4ALL is a nonprofit based out of Stanford University whose vision for AI focuses on building a pipeline for a diverse and inclusive workforce in AI, utilizing people with diverse backgrounds, voices and perspective to make better AI and making more tools for social good by redefining who can be a leader in AI. They have 3 programs: Changemakers in AI, AI4All Ignite, and their Open Learning curriculum. The AI4All Ignite internship program and Changemakers in AI are geared towards preparing undergraduate students from diverse communities for careers in AI. They provide mentors, support in technical interviews and internships as well as community support for students selected for the program. The Open Learning curriculum is foundational curriculum for high school students to help them learn about AI and how it works in various disciplines. They are focusing most of their energy on the college and career readiness programs, but their open learning curriculum is a solid starting place for any high school educator interested in getting their students started in AI regardless of subject area that is taught. The curriculum aligns to the following National Standards:
image source: https://AI-4-ALL.org There are lessons focused on explaining the more technical aspects of AI: "How Neural Networks Work", "How GANS Work", "How CNNs Work" and "How RNNs and Transformers Work". There are also lessons tied to specific topics and disciplines: "AI and Drawing", "AI and Facial Recognition" "AI and Deep Fakes: "AI and the Environment", "AI and Dance" and "AI Ethics". The lessons range from 1-10 hours long. Each lesson contains a detailed Teacher's Guide, Google Slide Deck, as well as a a study guide and google form for students to complete as they go through the lessons. The lessons are filled with experiential activities, explainer videos and discussion questions to allow students to grapple with the implications of how AI is changing our society. Each lesson also contains a spotlight on professionals from diverse communities who are involved in AI work related to the lesson. Most lessons have a project for students to complete once they have gone through the lesson. There is flexibility in how the lessons are taught and a thoughtful Online Strategy Guide and Discussion Strategy Guide provided in each lesson. They also recommend which subject areas each lesson can be taught in and provide relevant standards that are met with each lesson. It is a very thoughtfully crafted set of lessons on AI that intentionally provide detailed guidance so that those who might not otherwise feel comfortable teaching AI can quickly get a handle on AI topics to bring to their students.
image on freepik.com by vecstock I began learning about generative AI this past spring and since I was working with Rumie at the time I proposed making a microlearning course on generative AI. It was approved and I had a lot of fun creating the course which you can view here: Why is ChatGPT so popular? Learn about generative AI and how people use it. For an example of what generative AI is, I had ChatGPT create a poem about popsicles in the style of Jay-Z and then used Uberduck an audio AI program to generate an audio version of the poem in the voice of Jay-Z.
It has now been a year since ChatGPT came out and it is phenomenal to watch how generative AI is evolving and improving. I have been fascinated with its implications in the educational landscape. How will generative AI be used to personalize and differentiate learning? What policies will education systems come up with to use generative AI? How will assessments and learning experiences need to change now that students can easily access generative AI to create content? How do we amplify the benefits of AI while minimizing the risks in education? These questions have been forefront in my mind as I think about how to create Computer Science curriculum around AI. I watched a webinar this past summer entitled, Leveling up Digital Citizenship Skills with AI and it got me thinking about how we will have to teach students about AI in terms of responsible use and digital citizenship. The idea that students can get answers to homework or get ChatGPT to write an essay is understandably troubling for educators. The capabilities of generative AI are far more sophisticated than the era of being able to use google for an answer or essay and unfortunately for some adolescents the question is often not should I do it, but how do I not get caught at it. Teachers concern around this is absolutely legitimate. One worry I have is that it will make the work of teaching foundational skills like writing, math and even critical thinking increasingly challenging if educational systems don't get a handle on mitigating the risks with the easy access to ChatGPT by students. The way we need to engage and teach students is going to fundamentally change. The world of generative AI is truly going to demand we rethink education. The potential of AI to help not only students, but teachers do their work more efficiently is exciting and yet there is also so much to grapple with in regards to this innovation. I have also spent the last few weeks trying to really get a handle on the impacts of generative AI and how to best craft AI prompts and to really think about how it's potential for use in education. I found some really excellent free courses on Coursera through Vanderbilt University taught by Jules White as well as one by Google called "Introduction to Generative AI". The Google Course is a short 30 minute course that uses really clear graphics to highlight what is going on behind the scenes with generative AI. I completed two of White's courses called "Innovative Teaching with ChatGPT" and "Generative AI Primer". I highly recommend them for anyone interested in understanding how we need to think about using generative AI to amplify human creativity and problem-solving as well as how to engage ChatGPT through prompts to make the most out of it in teaching and other areas of our lives. I have also starting another course by White called "Prompt Engineering for ChatGPT" that I'm really excited about. The courses are lecture style, but White pulls up ChatGPT frequently and shows his process for writing prompts and how ChatGPT responds. One innovative feature of Coursera is that transcript notes can be found underneath the videos and you can highlight text to save as notes, but the really cool thing is that it not only saves the transcript notes, but also saves the video clip. I find it incredibly interesting and exciting to explore the challenges and innovations that generative AI brings to education. It has reinvigorated my passion for computer science education and digital citizenship and I hope I can find a role to be part of this reimagining. of the educational landscape.
Image Credit: https://yestem.org/tools/the-equity-compass/
One of the best professional development experiences I had this summer was taking part in an online course called, Equity in Informal STEM Learning: Using the Equity Compass created by University College London. The Equity Compass is powerful framework for assessing informal STEM learning such as after school clubs, summer camps and programs at museums.
I really appreciated the distinction they made between Equality described as “treating people in the same way, making sure people get the same opportunity” versus Equity described as “factoring in people’s different needs and assets, understanding that people might need different opportunities and support.” The Equity Compass focuses on four main parts: challenging the status quo, working with and valuing minoritized communities, embedding equity and extending equity. As participants of the course, we were introduced to each part of the framework and then presented with case studies to analyze that represented STEM Experiences that were lacking in terms of equity and diversity. We were also challenged to come up with solutions for each of the case studies to make them more equitable by applying our knowledge of each part of the framework. It was a well-designed course that really helped participants understand and apply practices to bringing more equity and diversity into informal STEM experiences. After taking the course, I reflected on the Equity Compass and the work Girls Who Code is doing with their Summer Immersion Program. I can proudly say that it is a powerful example of what informal STEM learning with a focus on equity and diversity looks like. It is truly an honor to be part of this work in getting more girls and students from marginalized communities involved in Computer Science in a way that values their experience. I recently completed a course on Adobe Captivate 2019 Fundamentals on Udemy. I also designed and created an elearning scenario-based training for parents called Screenwise Conversations. I really like Adobe Captivate even though Articulate Storyline is the most used software for elearning. If you are like and are used to the Adobe products and enjoy having a seamless workflow with those products. It's really advantage especially if you are a Mac user as I am and don't want to bog your computer down with a program like Parallels so that you can run PC only programs which unfortunately Articulate is. The two features that Adobe Captivate has that no other elearning software offers is responsive design with fluid boxes and also a really robust advanced actions feature that allows you to do really sophisticated interactions. I am still working on getting command of these features and chose to keep my elearning project simple. The course was great and creating my own elearning project while I took it really allowed me to hone and cement my skills with Adobe Captivate.
Since November, I have been volunteering as a learning experience designer for Rumie Learn. They specialize in creating Bytes which are microlearning courses that take 6-9 minutes to take and are aimed at social media users aged 14-29. The idea is to get people to scroll with purpose. Rumie has an excellent onboarding program where you go through a series of Bytes that are located together on Rumie Build, which is the content creation system used to create Bytes. Since each Byte is built in the format of the Bytes learning experience designers will be creating, you get a good sense of how to format a Byte while you getting the information and training you need to create Bytes. Brilliant! The learning director, Steve Birek, takes a very active role in supporting new learning experience designers and also giving support throughout the process with a weekly volunteer support Google Meet as well as quality feedback during the Byte creation process. Slack is used effectively to build teams or squads of designers and as a place to get support and get to know the Rumie Build Community. Rumie also uses Discord to connect the entire Rumie community connecting the learners who use Rumie with the designers and staff of Rumie. Microlearning course creation at Rumie involves a two-week Sprint structure. First, you choose a learning objective in Clickup, the project management software utilized throughout the Byte creation process. This learning objective will be the focus of your Byte. Then you have a week to design the first draft of your Byte. During the second week it goes through Peer Review where LXDs review each others' Bytes. Then the Byte is reviewed by a Byte Editor and finally published on Rumie Learn. I found it challenging and fun to work within the constraints of microlearning and also the Rumie Build system. The emphasis is on using clear, concise language and pictures, gifs, and memes to keep learners engaged. I chose to create Bytes on a wide range of topics. I have learned so much through this process and plan to continue to stay on as a learning experience designer to create more Bytes in the future.
|
AuthorYvonne Caples is a Learning Experience Designer who is passionate about making learning meaningful and engaging for all. Posts
All
|