TeachLab Presents The Homework Machine

We’re pleased to share powerful reporting from our friends at the show Click Here, from Recorded Future News and PRX. They tell true stories about the people making and breaking our digital world. Over the past year, they’ve been documenting the digital – and cultural – erasure of Uyghurs in China. As part of that series, they follow one man who set out to preserve his culture the old fashioned way: by opening kindergartens that celebrated the Uyghur language. And then, almost overnight, those doors were locked. And the school’s founder went from teacher… to enemy of the nation. Sean Powers reported the story and Dina Temple-Raston Dina, the host of the Click Here, brings it to us.It’s not too late to take our listener survey about the Homework Machine series. We will enter you in a drawing to receive one of two $25 gift cards.

The timing of the arrival of AI has been bad for many schools and teachers. In the Fall of 2022, schools were just emerging from the extraordinary challenges brought by the pandemic. Teacher morale and turnover are at historic lows. Some school districts are on a good footing, and have the resources and stability to adapt, and even experiment with using AI to support learning, but many teachers tell us generative AI has just been one more thing they have to deal with, often without the support they would like. At a moment when unprecedented sums are being invested in AI development — including many billions devoted to AI powered education technology — teachers wonder if our priorities are in the right place. This episode was produced by Steven Jackson and Jesse Dukes. We had editing from Ruxandra Guidi and Alexandra Salomon. Reporting and research from Holly McDede and Andrew Meriwether. Reporting and research for the series from Chris Bagg, Andrew Parsons, Natasha Esteves, and Marnette Federis. Sound design and music supervision by Steven Jackson.Production help from Yebu Ji. Data analysis from Manee Ngozi Nnamani and Manasa Kudumu. Administrative support from Jessica Rondon. Special thanks to Josh Sheldon and Eric Klopfer.Original music for this series was created by Steven Jackson, Andrew Meriwether and Jesse Dukes, as part of the music project Cue Shop. Thanks to Will Grueb, Andy Wilds, and the MIT Music Department for letting us use the MIT Harpsichord.The research and reporting you heard in this episode was supported by the Spencer Foundation, the Kapor Foundation, the Jameel World Education Lab, the Social and Ethical Responsibility of Computing initiative at MIT, and the RAISE initiative, Responsible AI for Social Empowerment and Education also at MIT. We had support from Google’s Academic Research Awards program.Please take our Listener Survey for a chance to win a $25 gift card.

https://docs.google.com/forms/d/15pFzNFcJ08xBSTnoE2-YJyoxsdIW5cKDnjrpBfBAf7U/edit

We keep saying that most ed tech is invited into schools, but generative AI crashed the party. But the truth in, some teachers have invited AI into schools. Some think it’s been great, and others see opportunities as well as challenges. We visit four classrooms where students and teachers are using AI in ways they say are innovative, fulfilling, and effective: a student in Oregon who worked with her teacher to expand the scope of her assignments with AI; a Southern California instructor using AI to bypass busywork and focus on the creative aspects of filmmaking; a teacher in Illinois deploying AI to provide targeted feedback; and, finally, a former engineer whose students use AI coding tools to make circuitboards…play Rick Astley? And we ask the question: What are the human teachers doing to create conditions in which AI can support learning? This episode was produced by Jesse Dukes. We had editing from Alexandra Salomon and Ruxandra Guidi. Reporting and research from Chris Bagg, and Andrew Parsons. Additional reporting from Andrew Meriwether, Holly McDede, Natasha Esteves, and Marnette Federis. Sound design and music supervision by Steven Jackson. Thanks to Jerry Shaw for digging up some of those Arduino Uno songs. The “Uptown Funk” was a simulation, created in our studios. Production help from Yebu Ji. Data analysis from Manee Ngozi Nnamani and Manasa Kudumu. Administrative support from Jessica Rondon. Special thanks to Josh Sheldon and Eric Klopfer. Original music for this series was created by Steven Jackson, Andrew Meriwether and Jesse Dukes, as part of the music project Cue Shop. Thanks to Will Grueb, Andy Wilds, and the MIT Music Department for letting us use the MIT Harpsichord.The research and reporting you heard in this episode was supported by the Spencer Foundation, the Kapor Foundation, the Jameel World Education Lab, the Social and Ethical Responsibility of Computing initiative at MIT, and the RAISE initiative, Responsible AI for Social Empowerment and Education also at MIT.We had support from Google’s Academic Research Awards program.

Education technology companies have been falling over themselves to provide AI powered tools for schools, and to convince schools to buy their product. The array of AI powered products is dizzying, and while some can likely be put to good use, it can be hard for school leaders to gauge which are worth adopting. One school system in Iowa, Winterset Community School District has gone all in on generative AI, piloting at least three AI powered products and encouraging faculty and students to experiment with AI. They kindly let us visit for a week, and we dive in to see what they’re excited about, and what has been challenging. This episode was produced by Andrew Meriwether and Jesse Dukes. We had editing from Ruxandra Guidi and Alexandra Salomon. Reporting and research for the Homework Machine from Natasha Esteves, Holly McDede, Andrew Parsons, Marnette Federis, and Chris Bagg. Sound design and music supervision by Steven Jackson. Production help from Yebu Ji. Data analysis from Manee Ngozi Nnamani and Manasa Kudumu. Special thanks to Josh Sheldon and Eric Klopfer. Administrative support from Jessica Rondon. Thanks to Chad Sussex and everybody at Winterset who spoke to Andrew Meriwether. We’re happy to share that AI Express: Leading the Future of Learning — written by Chad Sussex and Sandy Groom-Meeks, who guided the AI journey for Winterset — is now available. This book offers school leaders and educators a practical roadmap for safely and effectively implementing AI in schools. Find the book here: https://bit.ly/AIExpress_AmazonOriginal music for this series was created by Steven Jackson, Andrew Meriwether and Jesse Dukes, as part of the music project Cue Shop. Thanks to Will Grueb, Andy Wilds, and the MIT Music Department for letting us use the MIT Harpsichord. The research and reporting you heard in this episode was supported by the Spencer Foundation, the Kapor Foundation, the Jameel World Education Lab, the Social and Ethical Responsibility of Computing initiative at MIT, and the RAISE initiative, Responsible AI for Social Empowerment and Education also at MIT. We had support from Google’s Academic Research Awards program.

We’re halfway through “The Homework Machine” series, and as our production team works to finish the final episodes, we check in with Teaching Systems Lab director Justin Reich, who is returning to the classroom after a full year of sabbatical. Justin has been nervous, since some of the assignments he’s relied on in the past are not “AI-proof”. He tells us how he plans to adapt, We want to share A Guide to AI in Schools: Perspectives for the Perplexed the guidebook authored by TSL, based on the interviews we conducted for “The Homework Machine”. Justin is a fan of these protocols for evaluating student work:

We’re about halfway through The Homework Machine min-series, and we have more episodes coming. We're taking a break this week, which gives us the opportunity to share an episode of one of our favorite education podcasts with you.Mindshift from KQED features in depth interviews and reports from classroom about education and educators. We particularly enjoy this episode about a shift to teach contemporary poets, alongside the classics.Hanif Abduraqqib. Sarah Kay. Elizabeth Acevedo. Clint Smith. Do any of these names sound familiar? How about Amanda Gorman? All of these writers are part of America’s thriving contemporary poetry scene. But you won’t find them in many text books, because high school poetry units tend to focus on dead poets, like Robert Frost, Walt Whitman and Edgar Allen Poe. North Carolina teacher Melissa Smith is working to change that. For the last seven years, she’s been diversifying the canon in her classroom, and encouraging other teachers to do the same with the hashtag #teachlivingpoets. The shift has inspired teachers across the country to get creative with how they teach students things like tone, rhythm and structure in poetry. And it’s inspired students to connect with and see reflections of themselves in the poets they study.We have more episodes of the Homework Machine coming in a little over a week. In the meantime, please take our listener survey. (We'll enter you in a drawing to win a $25 gift card).https://forms.gle/KwPGTeVYZh2mo6gF7

Students today grapple with the allure of AI tools like ChatGPT for schoolwork, often blurring the lines of academic integrity. Amelia's struggle with a new English class leads her to rely on AI, impacting her relationship with her teacher. Meanwhile, high-achieving students face ethical dilemmas as they navigate the complexities of AI usage. As graduation approaches, the phenomenon of senioritis exacerbates cheating tendencies, highlighting the need for open dialogue about trust and support in education.

In this engaging discussion, Joe O'Hara, a high school chemistry teacher from Chicago, tackles the challenges of AI-assisted cheating in classrooms. He shares clever detection strategies, emphasizing the importance of evidence-gathering like screen monitoring. O'Hara explores the moral dilemmas teachers face between punishment and education, advocating for clear communication with parents to build trust. His innovative approach includes AI-proofing assignments, designing projects that promote ownership, and adapting to the evolving landscape of education.

ChatGPT is the most well known of the Large Language Models (LLMs) but what is an LLM? We go deep into how this remarkable new technology is built, and why their performance is inconsistent — or jagged — across similar tasks. We dive into the techniques AI engineers use to align these tools’ behavior with our values, and explain why they don’t always work, and sometimes we get hallucinations or biased output.  This episode was produced by Steven Jackson and Jesse DukesEditing:  Alexandra Salomon and Ruxandra Guidi  Reporting and research from Holly McDede, Natasha Esteves, Andrew Parsons, Andrew Meriwether, Marnette Federis, and Chris Bagg.Sound design and music supervision by Steven Jackson. Production assistance from Yebu Ji and Nathan Ray. Data analysis from Manee Ngozi Nnamani and Manasa Kudumu. Special thanks to Josh Sheldon, Camila Lee, Liz Hutner, and Eric Klopfer. Administrative support from Jessica Rondon. The research and reporting you heard in this episode was supported by the Spencer Foundation, the Kapor Foundation, the Jameel World Education Lab, the Social and Ethical Responsibility of Computing initiative at MIT, and the RAISE initiative, Responsible AI for Social Empowerment and Education also at MIT. We had support from Google’s  Academic Research Awards program. The Homework Machine is a program of the MIT Teaching Systems Lab, Justin Reich, director.