Education Futures

Kumar Garg is President of Renaissance Philanthropy, the science-and-technology philanthropy he co-founded with Tom Kalil in 2024 to help donors invest in high-impact research across AI, climate, and education. He holds a BA from Dartmouth and a JD from Yale Law School, and cut his teeth over eight years in the Obama White House Office of Science and Technology Policy, where he led the administration's STEM-education agenda — including the "Educate to Innovate" campaign that mobilized more than $1 billion in philanthropic and in-kind investment, and the White House Science Fair.

Before Renaissance, he spent nearly six years (2018–2024) as Vice President and Managing Director at Schmidt Futures, the philanthropic initiative of Eric and Wendy Schmidt, where he built much of its science, talent, and education portfolio and launched the Learning Engineering Virtual Institute (LEVI). He is one of the principal architects and evangelists of "learning engineering" — the idea that instruction should be treated as a systems-level engineering and data challenge — and delivered Carnegie Mellon's 2022 Simon Initiative Distinguished Lecture, "Shaping the Field of Learning Engineering." Today, Renaissance runs a DARPA-inspired "fund model" that raises capital against well-scoped three-to-five-year technical bets, from doubling middle-school math outcomes for low-income students to catching early-reading difficulties before third grade.

In this episode, Kumar talks with Svenia Busson about:

• The Renaissance "fund model" — how naming a bottleneck, recruiting a technical lead, and raising from donors like an investor raises from LPs unlocks three-to-five-year breakthroughs
• Why environmental factors like lead poisoning and classroom air quality may drive learning outcomes as much as pedagogy — and can move a child 10+ IQ points
• Learning engineering explained in plain language — the four always-running loops of learning science, design, instruction, and the individual learner
• LEVI Math's north star — doubling the rate of middle-school math gains for low-income kids at under $1,000 per student, down from the $4,000 high-dosage tutoring benchmark
• The LEVI Literacy Initiative — using AI as a diagnostic screener, not a teacher, to flag dyslexia and speech errors in the K–2 window
• "Dynamic dosing" and the human + AI tutoring triangle — why teachers and tutors enhance, rather than compete with, AI tools
• Building rigor into edtech — the Tools Competition's research-partner requirement and ending moonshots with published RCTs
• Why the advanced AI-and-education world is "shockingly small," and how DARPA-style active program management grows it

• Renaissance Philanthropy — https://www.renaissancephilanthropy.org/
• Learning Engineering Virtual Institute (LEVI) — https://learning-engineering-virtual-institute.org/
• LEVI Literacy Initiative — https://learning-engineering-virtual-institute.org/literacy/
• The Tools Competition — https://tools-competition.org/
• Big If True Science accelerator — https://www.renaissancephilanthropy.org/big-if-true-science-accelerator
• Schmidt Futures / Schmidt Sciences — https://www.schmidtsciences.org/
• Lead Exposure Elimination Project (LEEP) — https://leadelimination.org/
• Center for Global Development (lead pollution & learning research) — https://www.cgdev.org/
• J-PAL (2012 high-dosage tutoring research) — https://www.povertyactionlab.org/
• Carnegie Mellon "dynamic dosing" human + AI tutoring study (2025) — https://www.heinz.cmu.edu/media/2025/September/in-study-of-human-ai-tutoring-with-us-seventh-graders-human-tutors-enhanced-the-benefits-of-ai-tutors
• DARPA — https://www.darpa.mil/

Svenia Busson on building safe, pedagogically sound AI for young people and the fellowship Education Futures built to do it.

https://educationfutures.ai/fellowship

https://educationfutures.substack.com/

Maureen Russo Rodriguez is coming up on her 20th year as an educator and teaches language and literature — primarily Spanish, with occasional turns in English — at St. Mark's School, a small boarding school outside Boston. With a background in comparative literature and a stint in higher ed, she co-founded AI Co-Lab, a teacher-led, peer-to-peer network for building AI literacy. What began with a single faculty email she sent the week ChatGPT launched (a tool she shares a birthday with) has grown from nine teachers into a movement of 900+ educators across 300+ schools, with materials released under a Creative Commons license so any school can remix and reuse them.

Ron Spalletta teaches in the English department at the same school, with a background rooted in poetry (and a little molecular biology). Through a Patterson Innovation Grant, he has been probing what AI means for writing, learning, and assessment — and has designed several of Co-Lab's monthly "explorations," including a widely used one on alternative assessment built to help teachers profile themselves as assessors and rethink AI-resistant assignments.

In this episode, Maureen and Ron talk with Svenia Busson about:

• Defining AI literacy for teachers — awareness of what these tools are and aren't, in order to preserve the human core of teaching
• Why AI literacy is "a lot more driver's ed than automotive repair" — hands-on practice, not looking under the hood
• How Co-Lab works as a peer-to-peer network — structured monthly explorations, breakout rooms, and PD that pushes back on "one-and-done" webinars and consultant hype
• The four types of "naysayers" — and why refusing AI can be a legitimate, informed act of teacher agency
• Generative AI as an "arrival technology" — it's already in students' lives whether or not it enters your classroom
• Sarah Schneider Kavanagh's dilemma-and-decision cycle — how professional judgment breaks down if teachers never touch the tools
• Reinventing assessment in the age of AI — from oral exams to AI-resistant tasks, and pairing an LLM's scope with a teacher's discernment
• The students who worry them most — the ones doing everything right and watching peers coast on A-pluses through dishonest AI use

A conversation about why AI literacy isn't optional — it's how teachers keep their agency.

Links mentioned in the episode:

• AI Co-Lab: https://www.educolab.org/
• Justin Reich & Jesse Dukes, The Homework Machine podcast: https://www.teachlabpodcast.com/
• On "arrival technologies" (Reich & Dukes): https://www.researchgate.net/publication/385821042_Toward_a_New_Theory_of_Arrival_Technologies

Dr. Mathilde Cerioli is a cognitive neuroscientist working at the intersection of child development, AI, and public policy. She holds a Master's in child psychology from Paris and a Ph.D. in Cognitive Neuroscience from the Université de Montréal, where her research focused on using board games and video games to improve cognitive functioning in children and adolescents, including those with learning disabilities.

After her doctorate she worked with EdTech and video-game companies, translating child-development science into product design. She went on to co-found everyone.ai, a California-based nonprofit that anticipates, educates on, and evaluates the risks and opportunities of AI for children and youth (ages 0–25), alongside co-founders Grégory Renard, Anne-Sophie Seret, and Céline Malvoisin. As Chief Scientist, she leads its research, authored the founding report The Future of Child Development in the AI Era (2024), and recently launched the AïA Safety Builder — an assessment system that scores AI products on 15 human-like behaviors driving anthropomorphism in children. She is also Chief Scientist of the iRAISE Alliance, the global coalition launched with the Paris Peace Forum, and advises OpenAI's Expert Council on Well-Being and AI.

In this episode, Mathilde talks with Svenia Busson about:

• Why children don't have "mini adult brains" — and what developmental science says AI design keeps getting wrong
• The 15 human-like behaviors that make AI feel human — flattery, mimicry, validation, proactivity — and when they tip from helpful to harmful
• Anthropomorphism and attachment — why kids bond with machines faster than adults do
• Sensitive periods, pruning, and plasticity — how AI introduced too early risks cognitive "stunting" (echoing Rebecca Winthrop's work at Brookings)
• The adolescent brain, reward systems, and a rat-and-alcohol study that explains why teens crave AI approval
• Why friction and "expectation violation" in messy teen friendships aren't a bug — but how the brain actually learns
• Cognitive offloading in adults — metacognition, burnout, and the risk for young people entering the workforce
• Doing rigorous research when a new model ships every week — the iRAISE expert-consensus approach

If you are building AI tools in the space this is a must-listen.

Links mentioned:

• everyone.ai — https://everyone.ai
• iRAISE Alliance — https://i-raise.org
• Paris Peace Forum — https://parispeaceforum.org
• Rebecca Winthrop, Brookings Institution — https://www.brookings.edu/people/rebecca-winthrop/ and her article on cognitive stunting: https://www.brookings.edu/articles/is-it-time-to-measure-cognitive-stunting/

Henry Scott studies at the London Interdisciplinary School (lis.ac.uk), pursuing a Bachelor of Arts and Science in Interdisciplinary Problems and Methods, and will be part of the school's third graduating class. Dyslexic since childhood, he turned early to AI dictation tools to help articulate his thinking — a habit he credits with sharpening his own unassisted writing over time. At LIS he took an AI and Ethics module taught by legal scholar Emma Rengers, wrote a four-page AI declaration for a capstone project he partly "vibe-coded," and built a custom AI assistant — nicknamed Clover — to manage his calendar and track his mood and habits.

Felix Audebert is a first-year student at Forward College (forward-college.eu), the European university where students spend one year each in Lisbon, Paris, and Berlin. He is pursuing a double bachelor's in Politics & International Relations and Business & Leadership. Felix uses AI almost daily and has access to Forward College's AI hub, which includes a tool that flags how much of a student's work risks tipping into cognitive offloading. He recently used AI to help a Portuguese impact-measurement startup narrow down interview targets for a client project.

In this episode, Henry and Felix talk with Svenia Busson about:

• Using AI as an assistant rather than letting it think for you
• The line between cognitive offloading and freeing up mental bandwidth
• Why both still write by hand — and the NTNU brain-connectivity research (Frontiers in Psychology) behind it
• How Forward College and LIS each handle AI in the classroom, from AI hubs to mandatory declarations
• The honesty gap — why students under-report AI use, and tools like Consensus that can confirm your own bias back to you
• Rethinking assessment in the AI era, from written declarations to oral exams ("vivas")
• The EU AI Act (high-level summary), algorithmic bias in immigration and parole systems, and what it means for the workforce they're entering
• Emotional offloading — AI companionship, naming your assistant, and where human connection still wins

Oksana Matiiash is the CEO of WINWIN EdTech Center of Excellence, Ukraine's national hub for technology and innovation in education. Launched in December 2025 under Ukraine's Global Innovation Strategy, the Center generates evidence through research and foresight and pilots education technologies, including AI, while Ukraine navigates a full-scale war. It also brings together government, educators, researchers, startups, and global partners to scale innovations that improve learning. The Center's partners include the Ministries of Digital Transformation and of Education and Science of Ukraine, UNICEF, Google for Education, OpenAI, Estonian Centre for International Development (ESTDEV).

Oksana holds a Master of Public Policy from the University of Oxford, where she studied as a Weidenfeld-Hoffmann Scholar, and a law degree from Taras Shevchenko National University of Kyiv.

In this episode, Oksana talks with Svenia Busson about:

• Why Ukraine launched a national EdTech Center to accelerate education innovation during the war and in the age of AI
• Leading wartime emergency education — and what crisis taught her about what school systems actually need
• The WINWIN EdTech Center's three pillars: research and foresight, innovation and pilots, ecosystem-building
• Ukraine's new OpenAI-funded study on how AI is shaping learning for students living through years of disrupted schooling
• Why teachers, not students, come first in her AI rollout — even as kids learn to prompt on TikTok
• The World Bank-backed RCT where six weeks of tech-assisted tutoring closed months of learning loss
• Building a "Responsible AI" framework with teachers as co-designers, not just recipients
• The critical-thinking and learning-agency skills she considers non-negotiable in an information war

MacKenzie Price is the co-founder of Alpha School and the AI-driven 2 Hour Learning model that powers it, a Stanford-trained psychologist who left the traditional path the day her daughter came home from second grade and said, simply, "school is boring."

In 2014, working alongside the colleague she still calls "the OG, the Original Guide," Price opened the first Alpha campus in Austin, Texas, betting that AI-personalized, mastery-based instruction could compress a full day of core academics into two hours. The bet paid off: Alpha students now grow two to five-and-a-half times faster than peers on NWEA MAP Growth assessments, and the model has expanded to more than 50 campuses, including Texas Sports Academy (low-to-no-cost, athletics-focused, partly funded through Texas education savings accounts) and Waypoint Academy, a wilderness-based program. This fall she's taking the model international, with new World School campuses opening in Ecuador and Kenya, building on an earlier pilot that brought the model to roughly 1,000 Ukrainian refugee students.

Price built Alpha around three commitments: kids learn twice the material in a fraction of the time, kids build real-world life skills through afternoon workshops, and — first and foremost — kids love school. By Alpha's own surveys, 95% of students say they do, and as many as 60% say they'd rather be in class than on vacation. She hosts her own Future of Education podcast and shares Alpha's experiments with more than a million followers across social media.

In this episode, MacKenzie talks with Svenia Busson about:

• Why AI-driven mastery learning can compress a school day into two hours — without losing depth
• The "guide" model — trading lesson plans and lectures for one-on-one motivation and emotional coaching
• Cognitive load theory in practice — keeping every student in their own zone of productive struggle
• Paying kids to learn — the financial-literacy logic behind Alpha's motivation system
• Texas Sports Academy and the wilderness-based Waypoint Academy — turning "selection bias" into a design principle
• Bringing the model into public schools, and new World School campuses launching in Ecuador and Kenya
• Alpha X projects — the teenagers behind a self-produced Broadway musical, a six-figure jewelry brand, and CPR-certified rescue training
• Why MacKenzie believes philosophy, not computer science, is becoming education's most valuable subject

Jenny Anderson is an award-winning journalist and author with 25 years in the field, including a decade covering finance at The New York Times — where she won a Gerald Loeb Award in 2008 for her coverage of Merrill Lynch ahead of the 2008 financial crisis. She later pioneered coverage of the "science of learning" at Quartz, and now contributes to The New York Times, The Atlantic, and The Washington Post.

She is the co-author, with Rebecca Winthrop (Director of the Center for Universal Education at Brookings), of The Disengaged Teen: Helping Kids Learn Better, Feel Better, and Live Better (thedisengagedteen.com) — the product of five years of research, including a survey of more than 65,000 students and 2,000 parents conducted with Brookings and Transcend, into why so many children lose their love of learning in adolescence. She writes the weekly Substack newsletter How to Be Brave, reaching tens of thousands of educators and parents. She is a senior fellow at the Center for Teen Flourishing and co-host of Ask the Kids, a podcast with Transcend Education.

In this episode, Jenny talks with Svenia Busson about:

• The disengagement gap — why 75% of kids love school in primary years, but only 25% still do by 10th grade
• The Four Modes framework — Passenger, Achiever, Resistor, and Explorer — and why fewer than 4% of students land in Explorer mode
• Why phones and AI aren't the root cause of the teen mental-health crisis — academic pressure consistently ranks higher
• School models built for agency — Big Picture Learning's semester-long internships, Red Bridge School, and Valor Collegiate Academies' "school within a school"
• Assessment in the age of AI — competency-based learning, portraits of a graduate, and why parents resist abandoning high-stakes exams like the GCSEs
• The AI silence at home — why most teens use AI regularly while few talk to their parents about it
• AI, writing, and "cognitive stunting" — what outsourcing the first draft costs a developing thinker, building on Rebecca Winthrop's NYT piece on AI and creativity
• What parents can actually do — testing the tools themselves, and protecting space for productive struggle

Bethany Koby-Hirschmann is a designer, social entrepreneur, and co-founder and Chief Vision Officer of Fam Studio, a research and design practice based in Somerset, England. She holds a BA from the Rhode Island School of Design and an MSc in Responsibility and Business Practice from the University of Bath, and is completing a PhD on youth co-creation and the uses of enchantment. In 2012, after finding a discarded laptop in a skip near her home in East London, she co-founded Tech Will Save Us with her husband, Daniel Hirschmann, on the conviction that children should be producers of technology, not just consumers of it.

Tech Will Save Us grew into a STEAM company selling in 97 countries and partnered with the BBC, Microsoft, Samsung, and ARM to design the BBC micro:bit — a pocket-sized computer distributed free to a million UK children that has since reached more than four million users worldwide. After selling the company in 2021, Bethany founded Fam Studio, which co-creates with families and children to build technologies, learning content, and experiences centered on people and the planet. Current projects include a multimodal "Imagination Tool" that uses generative AI to bring children's voices into large-scale co-creation, and a wellbeing-and-AI research partnership with Oxford's Reuben College and its Child-Centred AI Design Lab.

In this episode, Bethany talks with Svenia Busson about:

• Finding a laptop in a skip — the origin story behind Tech Will Save Us
• Designing the BBC micro:bit with Microsoft, Samsung, and ARM — reaching four million children across 97 countries
• Why Fam Studio exists to serve "the village," not just the child
• Techno-optimism versus AI anxiety — holding both at once
• Whether AI reinforces the industrial-era school model or finally breaks it open
• Building the "Imagination Tool" — using generative AI to bring children's voices into co-creation at scale
• "No Tech Sundays" and the house rules her family set for her teenager's AI use
• Moving from human-centered to life-centered design — what biomimicry teaches educators

She closes with future-guest picks: Caroline Essame, author of Why Nature Matters (Routledge); Noan Fesnoux, creative adviser to Dubai's Museum of the Future (LinkedIn); Liz Robinson, CEO of Big Education (bigeducation.org); and Jenny Gibson of Cambridge's PEDAL Centre (pedalhub.net).

This special episode is a recording of a live panel discussion from the AI & Education Meetup series, hosted by Education Futures at the London Interdisciplinary School (LIS).

Svenia is joined by four guests at the intersection of AI and learning: Ash Brockwell, LIS associate professor and lead of the new Education Futures master's; Niccolò Pescetelli, LIS associate professor leading the AI & Collective Intelligence program; Stephen Jull, GeoGebra GmbH co-founder now Global Head of AI Educational Technology at Teach For All; and Bibi Groot, behavioral scientist and Chief Impact Officer at Eedi, who runs large-scale RCTs on AI tutors.

The conversation moves through what it means to "learn" when AI can retrieve and summarize anything instantly, the risk of cognitive offloading and "cognitive surrender," and evidence from a 12-week chess RCT on hints and productive struggle. The panel also digs into relationality and embodied learning, how Teach For All is building AI literacy at scale across 63 countries, tools designed to cut screen time while still using AI (like Eedi's QR-code diagnostic system), and closes with each guest's vision of a preferable — not just probable — future for education, including a sharp reminder from the audience that 70% of children globally still lack basic literacy access.

Recorded live in London as part of Education Futures' ongoing meetup series exploring desirable futures for education in the age of AI.