If you’re looking for PD services to help school administrators and teachers prepare students to enhance their learning using AI, you don’t have to search far to find AI for Education, based in New York. After reading her LinkedIn posts and perusing their website, I knew it was time to speak with Amanda Bickerstaff, the co-founder and CEO of AI for Education. Amanda is very clear about what AI for Education does. They help teachers and administrators learn to use Generative AI effectively, assist educational institutions in developing the “rules of the road” that will meet the needs of all school constituents, and help teachers integrate AI into their curricula and programs. Amanda has an ambitious goal, articulated in their mission:
“Provide AI literacy training to 1 million educators and lead the responsible adoption of Generative AI in the education ecosystem, empowering teachers and ultimately improving student outcomes while preparing them for the future.”
Her biggest challenge is keeping the company’s growth at pace with the growth and development of AI. According to Amanda, this is a unique challenge because “AI is the fastest-growing consumer-facing technology in history.”
Educate AI: Say more about the company’s mission.
Amanda: “At the foundational level, we provide AI literacy at scale for schools. The main focus is on leaders and teachers, primarily due to their openness and willingness to learn. As we grow, we will extend this to students as well. We aim to ensure that AI literacy is the foundation required to meet the challenges of technology and learning and transform the education system. While we don’t shout this from the rooftops, we understand that when AI literacy is established, it becomes apparent where the current structure is no longer suitable: traditional assessment is challenging to maintain in its existing format, innovation has been lacking, and supporting students has been a long-standing issue alongside foundational skill building, such as literacy. These issues were already stressed with the advent of devices and one-to-one computing. The Internet has been in place, but now its impact is accelerating.
AI literacy serves as a catalyst. It’s essential to think beyond merely learning what AI is and isn’t and to broaden our understanding to encompass the more practical goal of Generative AI literacy. We aren’t teaching about fine-tuning models or algorithmic thinking. Instead, we focus on Generative AI, which involves natural language—speaking or typing—to build knowledge with technology. That’s our objective. We often discuss what the future holds. We aim to participate in every stage of that evolution. After establishing our foundation over the past two years, we anticipate being part of the dialogue on AI literacy as it becomes increasingly recognized as a vital need for education. We want to engage in the conversation about redefining assessment. Our goal is to involve students in meaningful learning, aware that human skills will carry different significance in the future. Some practices we prioritize today may not hold as much importance tomorrow. This is an expanded version of our mission.”
Educate AI: What do you want to tell people that is not discussed on your website? [I had told Amanda that AI for Education’s website was comprehensive and concise. Its use of video brings to life much of what they are doing.]
Amanda: “We do have a foundation arm. We haven’t discussed it yet, but are excited about its potential. This means we will do much more work on AI literacy in marginalized communities.
What you can’t see on the website is the the overwhelmingly positive response we get to doing the work in our unique way. You’re an educator. How rarely do people enjoy PD? How often do they say, “This is the best PD I’ve ever participated in. I was scared of AI, and I’m not scared anymore. I’m going to try to shift my perspective.” This happens in 90 minutes to a day, which is pretty remarkable. So, I don’t think you can encapsulate the impact. I wish we could because we believe so strongly that the effect has to be sweeping and profound. It needs to be in the classroom. It needs to be in the staff room. It needs to be in the district offices. That’s the magic we want to happen. That’s what you can’t see on the website. I wish you could. It feels great.
I’m hosting our webinars. I monitor the chat, and it becomes apparent that everyone is kind, thoughtful, and sharing resources. We recently had 500 people in a webinar, and no one sounded resistant or disengaged, even in the face of some pretty complex stuff. We match the subject matter with the appropriate expertise, in this case, Eric Hudson (the webinar you watched on video while you were reviewing our site). You mentioned that you couldn’t see the chat, but I can tell you that many people were engaged in the conversation.
We’re fortunate. We have a strong community around us. Some people are uncomfortable with AI literacy being talked about in the same breath as schools and students because they fear what it will do to learning and their future as teachers, which is a profound concern for them. We’ll never be their favorites, but we keep a balanced approach and see ourselves as AI pragmatists. We’re not optimists or pessimists; we’re just focused on intentional, thoughtful integration and building trust in our ability to manage this new technology. That can be a surprise to some in light of our company name. It suggests that we are all in with AI for Education.”
Educate AI: I did PD several years ago when I was teaching, and you can tell when it’s clicking because it’s not a common occurrence. Much of the PD I did was required. These were topics that teachers didn’t choose to hear about in the first place as opposed to something that they attended because they were interested.
Amanda: “That’s what we see when participating in day-long workshops and conferences. If they start with a keynote, which people are willing to listen to, they can choose where they go for a breakout and frequently come to us. The room is often over capacity and feels very crowded. It gets exciting because we make people feel comfortable and safe, layered with a sense of purpose they now want to know. I think that is what has been helpful to our educators. I feel good about figuring out a model that makes people feel safe to experiment and ask questions. We create a space in which people aren’t so stuck in the conventional concerns, such as “get on the bus or get run over” or that they’re still talking and there’s going to be friction after the workshop ends because we’re leaving with some big buckets to fill and big rocks to move. Still, I don’t think we would participate in this chaos if we didn’t enjoy what we do.”
Educate AI: Where are we going with AI? What kind of journey will it be for us?
Amanda: “The first thing is that generative AI models’ cost and capacity are increasing significantly faster than other technological advancements. We’re not discussing Moore’s Law increases, doubling every four to 14 months. It’s much, much quicker. So, I think there needs to be a firm understanding that we can joke about being in the dial-up phase of Gen AI; it’s very early, only two and a half years. The experience is still pretty clunky right now, but what’s happening will be similar to the Industrial Revolution, except that it will happen in a decade, not a century. As a result, it will profoundly change how people work, communicate, learn, and experience medicine and diagnostics within the next three to five years. We will start seeing significant change when we recover from that whiplash. That’s when most tech devices will roll over. You’ll need a new computer; there’ll be grid and network changes and more capable tools. So that’s one area.
We will see this unfold in the two fields that will be most disrupted. One could argue that education is being disrupted, but I want to focus more on how AI influences the act of doing. Disruption is occurring right now at one level of education: the individual student and teacher level, particularly in middle and high school. However, the two areas experiencing the most disruption are rather unusual companions. You would never mention software engineering and writing in the same breath. If you assigned these skills to students, you might claim those two students couldn’t be more different. We will witness significant automation, job loss, and job redefinition in those fields, almost at a quantum rate of change. So I think that’s the second point.
When it comes to the technology itself, I have two thoughts. First, reasoning models, which can handle more significant tasks in the background based on simple prompts, will improve. Prompting still matters, and anyone who says otherwise is misleading you. Even with powerful reasoning models, how you interact with them and provide context for your prompts is crucial. However, we will see more reasoning models in the future. Second, I find the concept of agents intriguing; there’s real potential there. This suggests the ability to make part of our work autonomous. You might compare it to the low-tech scenarios of the past, like downloading a video overnight while you sleep. In the new realm of agents, you won’t need to sleep; they won’t take over your entire computer or consume all your internet bandwidth. It is appealing that a complex processing partner will respond to your priorities. It will do its job; when it’s ready, it will be like a toaster. The toast will pop up and say, “Here are the ten things you asked me to do.” It will present your lesson plan, but it will go further. It will also offer a thoroughly vetted unit plan, pulling in evidence-based research, aligning with departmental goals, and documenting a pedagogical approach. I think this will become increasingly common. I am also optimistic that the tools will become more reliable regarding access and output. Hopefully, there will be better fit-for-purpose tools in education.”
Educate AI: In the mid-90s, Frank Moretti worked with the Dalton School and Apple Education. He discussed “these new tools” as the wedge that would transform teaching and learning. However, it didn’t happen because we focused too much on the tools instead of the students. How much of a wedge will AI be for reinventing education and demanding that we pay attention to our students’ needs?
Amanda: “It’s going to be radical. We’re not talking about when the Internet was first made available to the general public and how long it took to impact schools. That timeframe spans decades. We’re looking at three to five years with AI. We’ll start seeing high-stakes assessments change out of necessity, including all the College Board exams and other significant nonprofit assessments. It’ll become increasingly difficult to distinguish whether someone has honed their prompting ability to the point where AI mimics a distinct voice and human-generated writing. Therefore, this will be a profound transformation. When will it happen? As I mentioned earlier, my estimate is three to five years. However, based on my experience with significant disruption, it may take eight years in education. In education’s defense, I must note that they are engaging in more professional development than other industries. Education for AI has already accomplished much of this. We’ve had a million visits to our website, and people are actively participating, so I believe there is significant potential to expand over the next two years.
Still, we recognize that the systems surrounding the competing priorities and needs of our schools and districts—many of which resemble a small city—hold schools back. This indicates that the focus cannot be solely on student learning, as doing so will limit the impact of AI. However, our high schoolers will encounter generative AI everywhere, whether they want to or not. At that point, the need for higher-order thinking skills and deeper learning becomes crucial before they head to college, where AI will undoubtedly play a role. When they reach college, consider a potentially disruptive business model: term papers, research papers, and dissertations can already be completed with the assistance of AI systems, provided there is effective prompting.”
Educate AI: Are you concerned that what will happen (again) is that all the leaky pipes in the bathroom will be fixed, but the bathroom will not be renovated?
Amanda: “Some places will advance quickly, and that’s where it gets scary. Who will leap ahead because of AI disruption? Will it occur in well-resourced communities, or will they be somewhat complacent due to their previous successes? Will it happen in the areas that need the most support? If it doesn’t, whether we’re focused on rural or urban education, they will fall even further behind. It’s a recurring question. That’s why I believe it is so important to apply what AI has to offer intentionally. I assure you that AI literacy will make a difference. It’s not like digital literacy, which examines how websites function and what’s factual and what’s not. You can ask AI for a response and then see it make mistakes, recognize that it’s random, or understand that it’s biased but can also be sycophantic. These issues don’t take long to address.”
Educate AI: I hope you will spend a few minutes reflecting on AI and writing. First, am I wrong to suggest that the plagiarism terror circulating among schools right now may be slightly overhyped? I admit that I was fortunate to work in a school where I had 12 kids in a classroom and knew their writing styles very well, but I can well imagine that if I had to grade 150 papers every time I assign writing, I wouldn’t be able to say much about those kids’ writing styles. So, in the latter setting, I could be fooled by ChatGPT, but that’s not as much an issue with AI as with the way we’ve traditionally taught writing.
Amanda: “Absolutely, a thousand percent. There are two points to consider. First, we know that traffic to the ChatGPT site decreases on Saturdays, Sundays, and during the summer. This suggests that students are utilizing it, leading me to conclude that it significantly impacts writing. We spoke with one of our AP English teachers, who is not much older than her students. She mentioned, “I’ve done everything to caution my students about using AI, and still, my students didn’t realize I could monitor their activity. They just went straight to ChatGPT, and the issues we discussed didn’t matter because the students didn’t see value in completing this assignment.” This leads to the idea that we’ve been assessing and teaching writing in ways that haven’t been meaningful to students for a long time. I fear students will not appreciate the skills involved until we can adequately change them to make writing more authentic.
The next two to three years could be challenging because kids are offloading. They’re using the Staples “that was easy” button. The easy button lets students say, “I don’t see the value in this assignment, or I am busy and don’t have time for work that has no meaning. And who can blame them? If Boston Consulting Group consultants were convinced, in five hours, to cut and paste 96% of an output that was sometimes wrong, then a high schooler in Queens, without those emotional reserves, those experiences, that raw knowledge of why we’re even doing this, will conclude the work is boring. So I think that’s the biggest challenge. I’m not sure that’s how teachers articulate the issue, but I think there is a fear that kids will stop learning meaningfully. It may be more of an issue for our ELA and English teachers that their students will not know how to write or will not do their writing assignments. And I think that is real. At least in the short term, it’s time to reassess how we do things, develop better materials, and create the actual time and support for shifting methodologies within the structures surrounding education. One teacher, school, or district will not be enough without that.
The writing issue is quite nuanced, which we explain to our educators. Technology may be binary at heart, but we are not. It doesn’t have to be a choice between writing the essay yourself or having ChatGPT write it for you. Schools and colleges spend considerable time defining the parameters for what can go on between the boundaries. That’s fine, but it doesn’t mean much to the students unless a rationale and context accompany the policy. Moreover, saying, “using AI in the writing process is plagiarism” is inaccurate.” This perspective lacks subtlety and nuance (lawyers might argue that’s a good thing). Instead, how about stating that you may use AI during writing? “We recognize that you may use it differently based on your strengths and weaknesses. There’s a significant difference between being unable to put pen to paper to correct one word here or there and altering the entire essay. But ChatGPT should simplify getting that word right. Is it acceptable to use AI to rephrase a phrase here or a sentence there, or perhaps to rewrite a paragraph I dislike after completing a draft on my own? Is that a valid approach? Then there’s the question of using AI to explore several steps deeper. Is that part of the assignment? Will students pursue it anyway if they are engaged in their topic and possess AI literacy?”
How do we get people to that space, able to articulate what AI is because it’s so non-intuitive that you cannot talk about GenAI without using it? You really can’t. And it’s not just a question of using it but correctly. There’s research to show that even the best model, if you ask it a basic prompt, will give you a basic answer. Excellent prompting skills are so impactful that it doesn’t matter what model of ChatGPT you’re using. You can use 3.5 (last year) with a good prompting technique and get almost as much, if not as much, as the current 4.5.
Our strategy centers on being well-informed and having current AI literacy. I am passionate about sharing our approach with as many educators and students as possible. Reimagining education will become a justifiable cause when we reach a critical mass.
About the interviewee
Amanda Bickerstaff is the Co-Founder and CEO of AI for Education. A former high school biology teacher and EdTech executive with over 20 years of experience in the education sector, she has a deep understanding of the challenges and opportunities that AI can offer. She is a frequent consultant, speaker, and writer on AI in education, leading workshops and professional learning across K12 and Higher Education. Amanda is committed to helping schools and teachers maximize their potential through the ethical and equitable adoption of AI.
About the interviewer
Joel Backon is the editor-in-chief of Educate AI Magazine. Before joining Educate AI, Joel launched and was the founding editor of Intrepid Ed News, an online education publication that grew to almost 100,000 readers. For most of his career, he was an independent schoolteacher and administrator, teaching history and government while leading the school’s instructional technology effort, running a dormitory, coaching basketball, and chairing several committees. Immediately following college, Joel spent 15 years in the printing and publishing industries.
