Published by XYZ, Date.
AI + Education Summit
Expand menus below, by clicking on arrows.
Introduction
September 30th, 2023
Betaworks: 29 Little W 12th St NYC
Organized by MeshEd, Betaworks, Emerson Collective + XQ
Can we trust AI with our kids?
Our symposium on AI + Education featured leading ed-tech designers, educational leaders, teachers, students of all ages, policy experts, and impact-oriented ed-tech investors.
We explored a series of curated demos of AI learning products that span early childhood to adult learning. We heard from thought leaders who facilitated discussion and gathered actionable input— demo by demo, using a real-time polling platform— on the merits and risks of deploying AI in the classroom.
This event was free for all and generously supported by the Emerson Collective and XQ Institute, who together with MeshEd and Betaworks, produced this symposium in order to break down some of the silos between ed-tech product developers, educators, policy makers, and students.
The goal of this event was to spark curiosity, encourage creative thought, and build an engaged community to positively influence the integration of AI into our educational systems.
The Symposium concluded with lunch and an optional afternoon collective debrief and reflection, in which we gathered student feedback and reflected on key findings from the morning.
What makes this symposium different?
- People. This symposium was an intentional mix of students (high school-grad school), educators, ethicists, policy-makers, AI builders, and investors.
- Companies. We invited a range of AI-for-learning companies including nonprofits, public benefit corporations, and startups at every stage of development.
- Data. We collected information from attendees before the event, during the event, and we will be distributing a post-event poll as well.
Key Speakers
Agenda
Demos
AI Product | Speaker | Website |
---|---|---|
Yasi Afsharnive, Satya Nitta | ||
Lili Davis, Simon S. | ||
Ian Serlin, Dave Levin, Aditi Garg, Mike Yates | ||
Michael Olaya | ||
Arman Jaffer | ||
Steven Johnson, Ann-Katherine Kimble | ||
Joseph South | ||
Amanda Bickerstaff | ||
Teryn Thomas | ||
Ferdinand Habsburg, Joseph Millogo | ||
Olga Batygin | ||
Courtland Leer, Vince Trost, Vineeth Voruganti |
All-Group Demos
We started the day with six all-group demos. Founders and developers had the opportunity to present their products to an audience of more than 150 people, including students, educators, AI builders, policy-makers, investors, and more.
- Merlyn Mind
- Khanmigo (Khan Academy)
- Playlab.ai
- Brisk Teaching
- MyLessonPal (Dexter)
- NotebookLMÂ (Google)
Fireside Chat
Four AI and Education experts joined us for a “Fireside Chat.” In this session, we discussed the ethical concerns of introducing AI into the classroom, the technological platforms upon which these products are built, and how to make AI safe for students under 18.
Speakers:
- Rob Reich
- Satya Nitta
- Liz O’Sullivan
- Vriti Saraf
Station-by-Station Demos
Following the all-group demos, we split into smaller groups for six station-by-station demos, where audience members had the opportunity for close-up interaction with founders and their products.
- AI for Education
- Stretch AI(ISTE)
- EdLight
- EdCompass
- Lucinetic
- Bloom (Plastic Labs)
Data
For the AI + Education Summit, we used AhaSlides to create a more interactive conference experience—and to gather participants input and questions in realtime. We polled the audience in real time to capture their excitement and concern with regard to AI’s potential impact on learning before and after hearing the demos, as well as their reactions to and questions about the demos. Audience members generally expressed a large degree of optimism, and varied degrees of concern, which remained fairly steady from the pre-demo poll to the post-demo poll.
**The data collected from the poll represents only those audience members who chose to respond. There was a significant drop in responses from the pre-demo polls (~81% participation) to the post-demo polls (~50% participation). We believe this could be due to audience fatigue, as well as the amount of time we allowed for the audience to respond. We welcome your feedback on how to improve our data collection for next year.
Findings:
- The top 3 categories participants were most optimistic about with regard to AI’s potential impact on learning were: learning experience personalization, automation of mundane teacher tasks, and ongoing assistance during learning.
- The top 3 categories participants were most concerned about with regard to AI’s potential impact on learning were: algorithmic bias, equitable access to the best AI tools, and student and teacher data privacy.
Pre-Demo General Polls
**This data represents ~81% audience participation.
Post-Demo General Polls
**This data represents ~50% audience participation.
Demo Reaction Polls
**This data represents only a small portion of the audience, ranging from 16-71 responses (10%-46% participation) for each demo. This is likely because we had limited time for polling between demos.
Merlyn Mind:
Playlab:
Brisk Teaching:
Khanmigo:
MyLessonPal:
NotebookLM:
Post-Event Survey
We will keep updating this section as more responses come in to our post-event survey (link to survey HERE). Here are a few highlights from the 5 responses that have come in so far.
Overall Success of Symposium
Comments + Feedback
Have your interests in AI in Education changed since attending the event? Please explain.
- “Yes. As someone who is working on a shift to Standards Based Grading in our district, some of the demos and their ability to support rubric development throughout AI were impressive.”
- “Not directly related to the interests above; but I'd like to see more access provided to public school settings and historically marginalized populations!”
Have your concerns in AI in Education changed since attending the event? Please explain.
- “Many of the tools are focused on enhancing the experience of the 'teacher' -- other than NotebookLM or Khanmigo, I'd like to see more tools that enhance the experience of a high school student”
- “I think that most of my concern is around the development of AI and the intentional uses of it in combination with all the other AI tools.”
Additional comments and suggestions:
- “Would have loved more time/structure for making connections with other attendees.”
- “The best part of the event was making connections with amazing and inspiring people I had not met before”
- “Some operational and logistical details/signage would have been helpful (i.e. name tags, wifi password, bathroom, breakouts by role).”
- “Thank you for the invitation and elevating student voice in the process.”
- “The presentations at the end---where we moved among six stations---was difficult: hard to hear, groups were too big, unclear where to go and when rotations ended.”
Polling
We experimented with AhaSlides for the first time in an effort to create an interactive conference experience and collect real-time data. We received helpful feedback to use this tool more effectively in the future:
- “It was helpful; but please go back to it and maybe elicit some participant input to elaborate on their responses”
- “went too fast, hard to juggle between note taking, looking at websites of presenters, and polling page”
- “I thought that it was interesting, but it wasn't facilitated super well, so I am worried about the validity of the data.”
Attendee Key Takeaways
Credly Digital Badges from the New York Academy of Sciences for All Attendees
All participants will receive a badge from NYAS on AI and Education. The badge will be issued directly from Credly to the email that you used to register.
Teacher PD Credits (CTLEs)
We are offering professional development credentials to all New York State educators who attended the symposium thanks to our generous partner, The New York Academy of Sciences. Participants will receive a completion certificate from the New York Academy of Sciences for seven continuing education professional development hours (CTLEs).
Core Research on AI + Education
- Click here for list of readings, or download below.
Demo Links
Available now. See links below to sign up or explore free online products.
Limited availability. See links below to request early access.
- Playlab.ai
- NotebookLM (All attendees with gmail accounts will receive early access to NotebookLM)
- EdCompass
Not yet available to the market. See links below for more information.
High School AI Leaders Program
The High School AI Leaders program is a team of 20 high school seniors elected by their schools to help us investigate the ethical and practical questions relating to the use of AI tools in education. Each student attended two pre-event Zoom coaching meetings, they read through excerpts from THIS LIST of research documents and articles on AI for eduction, and they attended Day 1 of the Symposium. At the Symposium, students observed and asked each AI tool presenter two pre-prepared topical questions to explore each AI tool’s implications for the future of learning.
The vision of this program was to give high school seniors the opportunity to engage with the latest Ed Tech tools and empower their voices in this rapidly developing learning landscape. This program connects directly to the mission of this symposium, which is to break down silos separating educators, students, tech developers and policy makers and to foster partnerships and collaboration among key stakeholders in the educational ecosystem.
Pre-Event: The High School AI Leaders met for two pre-event meetings with program facilitators and undergraduate student coaches. Students completed a set of readings to contextualize their understanding of AI innovations in education and the work of the companies set to demo their AI tools at the event. Students were paired with two demoing companies and an undergraduate student coach, who supported them in crafting topical questions to ask of their assigned demos. Student questions focused on the practical use, ethics, and implementation the demoed AI tools.
The Event: The student AI Leaders joined world class leaders in computer science, educational design, and school leadership to view demonstrations and hear from thought leaders about AI tools for learning. The morning sessions featured six all-group demos. During a Q&A period following each demo, student AI Leaders asked the presenters their prepared topical questions. One star student, Simon, did a live demonstration of Khanmigo, Khan Academy’s new AI tutoring chatbot. Following his demo, Simon and his on-stage facilitators engaged the audience in a discussion of the practical use-cases for this tool. The afternoon session featured station-by-station demos. AI Student Leaders asked questions following each demo, utilizing the opportunity to engage presenters in a small-group setting.
Once demos concluded, student AI Leaders, their coaches, and many of the demoing presenters joined in a debrief and reflection. See below in the Optional Collective Debrief and Reflection section for more details.
Post-Event: Students will reconvene with coaches and program facilitators for an optional meeting on October 27th. Here is what the students take away from the symposium:
- Students will be guided through a resume building workshop and supported in finding internships in their fields of interest.
- Students will be given their Credly digital badge as a certification of their event participation.
- Students will be given a list of AI for Education companies to which they can apply for paid summer internships.
You can read all students’ topical questions asked during the event by clicking in here
Optional Collective Debrief and Reflection
In our collective debrief and reflection, we considered the data we gathered from 1) our pre-event poll (100 responses) and 2) our real-time polling (average: 73 responses) as well as additional prompts. This was primarily an activity with the High School AI Leaders and their coaches, but we welcomed participation from anyone who was interested.
The group analyzed data from the pre-event poll, as well as the real-time poll data (see above Polling section for more details) to gauge collective responses to the demos. Students then shifted into a reflection of the event as a whole, and a self-reflection regarding their experiences.
Highlights from group discussion:
- Students noticed some categories were frequently addressed by demos (learning experience personalization, automation of mundane teacher tasks, and ongoing assistance during learning), while other categories were seldom addressed (administrative workflows).
- Students expressed a desire to see more AI products catered to special needs students.
- Students questioned the financial accessibility of these tools for low income schools and students.
- Students brought up concern with AI algorithmic bias, particularly with regard to chatbots.
Highlights from student reflection notecards:
- Something that excites me is…
- “the possibility for more universal knowledge.”
- “startups are targeting the problems educators care the most about.”
- “the possibilities of actually learning at your own pace and own level.”
- Something that concerns me is…
- “how easy and fast AI tools are. They might replace people’s jobs.”
- “the algorithmic bias Khanmigo holds when solving math problems.”
- “the opportunity that some AI tools give students to be academically dishonest. Regulation is needed to ensure students are properly using these tools.”
You can view student reflection notecards here.