LearnLM is our new family of models fine-tuned for learning, and grounded in educational research to make teaching and learning experiences more active, personal and engaging.
We often talk about what Generative AI will do for coders, healthcare, science or even finance, but what about the benefits for the next generation? Permit me if you will, here I’m thinking about teachers and students.
It’s no secret that some of the most active users of ChatGPT in its heyday, were students. But how are other major tech firms thinking about this?
I actually think one of the best products with the highest ceiling from Google I/O 2024 is LearnLM. It has to be way more than a chatbot, it has to feel like a multimodal tutor. I can imagine frontier model agents (H) doing this fairly well.
What if everyone, everywhere could have their own personal AI tutor, on any topic?
ChatGPT4o Is the TikTok of AI Models — from nickpotkalitsky.substack.com by Nick Potkalitsky In Search of Better Tools for AI Access in K-12 Classrooms
Nick makes the case that we should pause on the use of OpenAI in the classrooms:
In light of these observations, it’s clear that we must pause and rethink the use of OpenAI products in our classrooms, except for rare cases where accessibility needs demand it. The rapid consumerization of AI, epitomized by GPT4o’s transformation into an AI salesperson, calls for caution.
[On 5/21/24] at Microsoft Build, Microsoft and Khan Academy announced a new partnership that aims to bring these time-saving and lesson-enhancing AI tools to millions of educators. By donating access to Azure AI-optimized infrastructure, Microsoft is enabling Khan Academy to offer all K-12 educators in the U.S. free access to the pilot of Khanmigo for Teachers, which will now be powered by Azure OpenAI Service.
The two companies will also collaborate to explore opportunities to improve AI tools for math tutoring in an affordable, scalable and adaptable way with a new version of Phi-3, a family of small language models (SLMs) developed by Microsoft.
Khan Academy’s AI assistant, Khanmigo, has earned praise for helping students to understand and practice everything from math to English, but it can also help teachers devise lesson plans, formulate questions about assigned readings, and even generate reading passages appropriate for students at different levels. More than just a chatbot, the software offers specific AI-powered tools for generating quizzes and assignment instructions, drafting lesson plans, and formulating letters of recommendation.
…
Having a virtual teaching assistant is especially valuable in light of recent research from the RAND Corporation that found teachers work longer hours than most working adults, which includes administrative and prep work outside the classroom.
The fact that a district could struggle so mightily with special education staffing that students are missing school – that’s not just a Del Norte problem. A recent federal survey of school districts across the U.S. found special education jobs were among the hardest to staff – and vacancies were widespread. But what’s happening in Del Norte is extreme. Which is why the Lenovers and five other families are suing the school district, as well as state education leadership, with help from the Disability Rights Education and Defense Fund.
…
The district sits hidden away like a secret between Oregon, the frigid Pacific and some of the largest redwood trees in the world. It’s too isolated and the pay is not competitive enough, Harris says, to attract workers from outside Del Norte. Locally, these aides – like the one Emma requires – earn about as much as they would working at McDonald’s.
From DSC: My wife does a lot of work with foster families and CASA kids, and she recommends these resources for helping children who have experienced adversity, early harm, toxic stress, and/or trauma.
TBRI® is an attachment-based, trauma-informed intervention that is designed to meet the complex needs of vulnerable children. TBRI® uses Empowering Principles to address physical needs, Connecting Principles for attachment needs, and Correcting Principles to disarm fear-based behaviors. While the intervention is based on years of attachment, sensory processing, and neuroscience research, the heartbeat of TBRI® is connection.
The adoption of a child is always a joyous moment in the life of a family. Some adoptions, though, present unique challenges. Welcoming these children into your family–and addressing their special needs–requires care, consideration, and compassion. Written by two research psychologists specializing in adoption and attachment, The Connected Child will help you:
Build bonds of affection and trust with your adopted child
Effectively deal with any learning or behavioral disorders
Discipline your child with love without making him or her feel threatened
Over the past 120 days, we’ve conducted tours of over 50 high schools in more than 1,000 classrooms across various cities including Boston, Dallas, Los Angeles, Northern Colorado, Kansas City, Twin Cities, Pittsburgh, and San Diego. These schools were purposefully selected for their dedication to real world learning, positioning them at the forefront of innovative education. These visits showed schools leading the way into new pathways, active learning methods, and work-based learning initiatives. From our observations at these leading schools, we’ve identified 8 key insights about the state of American high schools.
We are on the brink of a significant transformation in how education qualifications are perceived and valued, thanks to a strategic move by ETS to make Mastery Transcript Consortium (MTC) a subsidiary. This pivotal development marks a shift from traditional metrics of educational success—courses and grades—to a more nuanced representation of student abilities through skills transcripts.
The partnership between ETS and MTC is not just a merger of organizations, but a fusion of visions that aim to recalibrate educational assessment. The collaboration is set to advance “Skills for the Future,” focusing on authentic, dynamic assessment methods that provide clear, actionable insights into student capabilities. This shift away from the century-old Carnegie Unit model, which measures educational attainment by time rather than skill mastery, aims to foster learning environments that prioritize personal growth over time spent in a classroom.
As we move forward, this approach could redefine success in education, making learning experiences more adaptive, equitable, and aligned with the demands of the modern world.
One of the core problems is that education is based on time rather than learning.
We finally have a chance to move courses and grades into the background and foreground powerful personalized learning experiences and capture and communicate the resulting capabilities in much more descriptive ways—and do it at scale
Kockler hypothesizes that the reading struggles of many older students are due in large part to two issues. One has to do with “linguistic difference.” If a child’s family and community speak a variant of English that differs from the kind generally used in books and by teachers—for example, African-American English—it could be harder for them to decode words and connect those words to their meanings.
The Decoding Threshold The other issue has to do with difficulties in decoding multisyllabic words.Kockler points to a couple of large-scale research studies that have identified a “decoding threshold.”
In theory, students’ reading comprehension ability should improve as they advance to higher grade levels—and it often does. But the researchers found that if students are above fourth grade—past the point where they’re likely to get decoding instruction—and their decoding ability is below a certain level, they’re “extremely unlikely [to] make significant progress in reading comprehension in the following years.” The studies, which were conducted in a high-poverty, largely African-American district, found that almost 40% of fifth-graders and 20% of tenth-graders included in the sample fell below the decoding threshold.
The education profession relies on teachers being accessible to their students and families and open to sharing with colleagues. But a little information can be a dangerous thing.
In 2019, prior to remote learning, approximately 2.5 million students were homeschooled in the United States. This number has risen significantly, with estimates indicating that almost 4 million students are being homeschooled nationwide.
The homeschooling educational approach allows parents or guardians to educate their children at home rather than sending them to a traditional public or private school. Homeschooling families typically design their own curriculum or use pre-designed curricula tailored to their children’s needs and interests. In many cases, homeschooling is a choice made by families seeking more flexibility, personalized instruction, or alignment with their values and beliefs about education
Hello GPT-4o — from openai.com We’re announcing GPT-4o, our new flagship model that can reason across audio, vision, and text in real time.
GPT-4o (“o” for “omni”) is a step towards much more natural human-computer interaction—it accepts as input any combination of text, audio, image, and video and generates any combination of text, audio, and image outputs. It can respond to audio inputs in as little as 232 milliseconds, with an average of 320 milliseconds, which is similar to human response time in a conversation. It matches GPT-4 Turbo performance on text in English and code, with significant improvement on text in non-English languages, while also being much faster and 50% cheaper in the API. GPT-4o is especially better at vision and audio understanding compared to existing models.
Providing inflection, emotions, and a human-like voice
Understanding what the camera is looking at and integrating it into the AI’s responses
Providing customer service
With GPT-4o, we trained a single new model end-to-end across text, vision, and audio, meaning that all inputs and outputs are processed by the same neural network. Because GPT-4o is our first model combining all of these modalities, we are still just scratching the surface of exploring what the model can do and its limitations.
This demo is insane.
A student shares their iPad screen with the new ChatGPT + GPT-4o, and the AI speaks with them and helps them learn in *realtime*.
Imagine giving this to every student in the world.
Learning to Work, Or Working to Learn? — from insidehighered.com by Erin Crisp; via Melanie Booth, Ed.D. on LinkedIn We need a systems approach to making work-to-learn models just as accessible as traditional learn-to-work pathways, Erin Crisp writes.
Over the past two years, I have had the unique experience of scaling support for a statewide registered teacher-apprenticeship program while also parenting three college-aged sons. The declining appeal of postsecondary education, especially among young men, is evident at my dinner table, in my office, and in my dreams (literally).
…
Scaling a statewide apprenticeship program for the preparation of teachers has meant that I am consistently hearing from four stakeholder groups—K-12 school district leaders, college and university leaders, aspiring young educators, and local workforce development leaders.
A theme has emerged from my professional life, one that echoes the dinner table conversations happening in my personal life: Society needs systematic work-to-learn pathways in addition to the current learn-to-work ecosystem. This is not an either/or. What we need is a systematic expansion of effort.
In a work-to-learn model, the traditional college sequence is flipped. Instead of starting with general education coursework or survey courses, the working learner is actively engaged in practicing the skills they are interested in acquiring. A workplace supervisor often helps him make connections between the coursework and the job. The learner’s attention is piqued. The learning is relevant. The learner gains confidence, and seeing their influence in the workplace (and paycheck) is satisfying. All of the ARCS model elements are easily achieved.
I quickly decided to take a different tack with my students, and instead asked each of them, “What problem in the world do you think you want to solve? If you could go to a school of hunger, poverty, Alzheimer’s disease, mental health … what kind of school would you want to attend?” This is when they started nodding vigorously.
What each of them identified was a grand challenge, or what Stanford d.school Executive Director Sarah Stein Greenberg has called: purpose learning. In a great talk for Wired, Greenberg asks,
What if students declared missions not majors? Or even better, what if they applied to the School of Hunger or the School of Renewable Energy? These are real problems that society doesn’t have answers to yet. Wouldn’t that fuel their studies with some degree of urgency and meaning and real purpose that they don’t yet have today?
However, in my experience, far too often there is too little time actually spent considering and reflecting upon the choices we make as teachers.
So that’s what I wanted to write about today—along with a tool that I’m using refocus my own reflections around the choices I’ve made in my classroom this year. .
How do I put positive behavior strategies into practice?
How can families support this at home?
Additional resources and research
Empathetic sentence starters for teachers — from understood.org by Amanda Morin Learn how sentence starters can help you respond to your students with empathy. Get a one-page printable of empathetic sentence starters to help you get started.
Before jumping into fix-it mode, ask some open-ended questions and give kids time to answer. Try not to make assumptions or impose your viewpoint. Use sentence starters like these:
For the first time ever, the average teacher salary in America has topped $70,000 this school year.
Teacher salaries are growing faster than at any previous point since the Great Recession—but with inflation factored in, teachers still take home less on average than they did a decade ago, according to the nation’s largest teachers’ union.
the internet eliminated time and place as barriers to education, and
generative AI eliminates access to expertise as a barrier to education.
Just as instructional designs had to be updated to account for all the changes in affordances of online learning, they will need to be dramatically updated again to account for the new affordances of generative AI.
The Curious Educator’s Guide to AI | Strategies and Exercises for Meaningful Use in Higher Ed — from ecampusontario.pressbooks.pub by Kyle Mackie and Erin Aspenlieder; via Stephen Downes
This guide is designed to help educators and researchers better understand the evolving role of Artificial Intelligence (AI) in higher education. This openly-licensed resource contains strategies and exercises to help foster an understanding of AI’s potential benefits and challenges. We start with a foundational approach, providing you with prompts on aligning AI with your curiosities and goals.
The middle section of this guide encourages you to explore AI tools and offers some insights into potential applications in teaching and research. Along with exposure to the tools, we’ll discuss when and how to effectively build AI into your practice.
The final section of this guide includes strategies for evaluating and reflecting on your use of AI. Throughout, we aim to promote use that is effective, responsible, and aligned with your educational objectives. We hope this resource will be a helpful guide in making informed and strategic decisions about using AI-powered tools to enhance teaching and learning and research.
Annual Provosts’ Survey Shows Need for AI Policies, Worries Over Campus Speech — from insidehighered.com by Ryan Quinn Many institutions are not yet prepared to help their faculty members and students navigate artificial intelligence. That’s just one of multiple findings from Inside Higher Ed’s annual survey of chief academic officers.
Only about one in seven provosts said their colleges or universities had reviewed the curriculum to ensure it will prepare students for AI in their careers. Thuswaldner said that number needs to rise. “AI is here to stay, and we cannot put our heads in the sand,” he said. “Our world will be completely dominated by AI and, at this point, we ain’t seen nothing yet.”
Is GenAI in education more of a Blackberry or iPhone? — from futureofbeinghuman.com by Andrew Maynard There’s been a rush to incorporate generative AI into every aspect of education, from K-12 to university courses. But is the technology mature enough to support the tools that rely on it?
In other words, it’s going to mean investing in concepts, not products.
This, to me, is at the heart of an “iPhone mindset” as opposed to a “Blackberry mindset” when it comes to AI in education — an approach that avoids hard wiring in constantly changing technologies, and that builds experimentation and innovation into the very DNA of learning.
…
For all my concerns here though, maybe there is something to being inspired by the Blackberry/iPhone analogy — not as a playbook for developing and using AI in education, but as a mindset that embraces innovation while avoiding becoming locked in to apps that are detrimentally unreliable and that ultimately lead to dead ends.
Randomized-controlled experiments investigating novice and experienced teachers’ ability to identify AI-generated texts.
Generative AI can simulate student essay writing in a way that is undetectable for teachers.
Teachers are overconfident in their source identification.
AI-generated essays tend to be assessed more positively than student-written texts.
Can Using a Grammar Checker Set Off AI-Detection Software? — from edsurge.com by Jeffrey R. Young A college student says she was falsely accused of cheating, and her story has gone viral. Where is the line between acceptable help and cheating with AI?
ChatGPT shaming is a thing – and it shouldn’t be — from futureofbeinghuman.com by Andrew Maynard There’s a growing tension between early and creative adopters of text based generative AI and those who equate its use with cheating. And when this leads to shaming, it’s a problem.
Excerpt (emphasis DSC):
This will sound familiar to anyone who’s incorporating generative AI into their professional workflows. But there are still many people who haven’t used apps like ChatGPT, are largely unaware of what they do, and are suspicious of them. And yet they’ve nevertheless developed strong opinions around how they should and should not be used.
From DSC: Yes…that sounds like how many faculty members viewed online learning, even though they had never taught online before.
Last year at the Kalamazoo Valley Museum, kids could learn about how misinformation is made and how to avoid it. Now the media scholar behind the exhibit is adapting it for libraries.
A new exhibit for libraries
That concern also drove the “Wonder Media” exhibit that ran through last year at the Kalamazoo Valley Museum. Sue Ellen Christian is a communications professor at Western Michigan University. The exhibit was her idea. Full disclosure: I’m a former student of Christian’s. We met in her office on campus.
“It’s really important for our entire society to think about the importance of facts and truth to a democracy,” said Christian. “And without an informed citizenry, we cannot have a healthy democracy.”
Christian recently received a grant from the Institute of Museum and Library Services, based in Washington D.C., to adapt the Wonder Media exhibit for public libraries. It’s designed to reach middle-school-age children.
Mainly, with her grant, Christian wants to develop something for students whose schools do not have librarians anymore. The website associated with the exhibit has resources for students, teachers, and libraries.
OpenAI rolls out Memory feature for ChatGPT
OpenAI has introduced a cool update for ChatGPT (rolling out to paid and free users – but not in the EU or Korea), enabling the AI to remember user-specific details across sessions. This memory feature enhances personalization and efficiency, making your interactions with ChatGPT more relevant and engaging.
.
Key Features
Automatic Memory Tracking
ChatGPT now automatically records information from your interactions such as preferences, interests, and plans. This allows the AI to refine its responses over time, making each conversation increasingly tailored to you.
Enhanced Personalization
The more you interact with ChatGPT, the better it understands your needs and adapts its responses accordingly. This personalization improves the relevance and efficiency of your interactions, whether you’re asking for daily tasks or discussing complex topics.
Memory Management Options
You have full control over this feature. You can view what information is stored, toggle the memory on or off, and delete specific data or all memory entries, ensuring your privacy and preferences are respected.
Memory is now available to all ChatGPT Plus users. Using Memory is easy: just start a new chat and tell ChatGPT anything you’d like it to remember.
Memory can be turned on or off in settings and is not currently available in Europe or Korea. Team, Enterprise, and GPTs to come. pic.twitter.com/mlt9vyYeMK
From DSC: The ability of AI-based applications to remember things about us will have major and positive ramifications for us when we think about learning-related applications of AI.
The model, known as team teaching, isn’t new. It dates back to the 1960s. But Arizona State University resurrected the approach, in which teachers share large groups of students, as a way to rebrand the teaching profession and make it more appealing to prospective educators.
…
In Mesa, teachers working on a team leave their profession at lower rates, receive higher evaluations and are more likely to recommend teaching to a friend.
Early research also indicates students assigned to educator teams made more growth in reading and passed Algebra I at higher rates than their peers.
It is tough to bring students back to the classroom once chronic absenteeism rates begin to climb. As more districts struggle with historically high absenteeism, new research suggests they may need a more systemic approach to reengaging students.
A new working paper on Michigan schools released by the Annenberg Center found most school districts with severe attendance problems did not directly address absenteeism when planning school improvement strategies. Among those that did focus on improving attendance, few coordinated their interventions across schools and aligned interventions to combat the specific barriers keeping students from school.
“If you think about the reasons that families are missing school, informing families about their children’s attendance is certainly important, but it’s not like the primary driver of absenteeism,” Singer said, “so there’s a disconnect.”
However, I’ve learned that effective leadership isn’t about having all of the answers; it’s about knowing which questions to ask. Effective leaders listen deeply and ask questions that shape people’s thinking, moving the organization from where it is to where it needs to go.
Today’s K12 students are spending the vast majority of their time in classrooms listening to answers to questions they did not ask and following rules they did not have a hand in making. Given that this dynamic goes on for years, what is it doing to students’ minds and spirits? To their agency and empowerment? Are we unintentionally graduating dependent young adults?
…
But what if the opposite were true? What if schools empowered children to flourish? What if schools were the places where they could explore, identify, express, and develop their thoughts, feelings, and goals? There’s power in the uniqueness of every child. It’s time that school designs honor students’ unique calling, preferences, and goals, and encourage them to pursue those. It’s time to move fully into a new era for learning where learners can develop greater self-leadership than ever before.
