AI Pedagogy Project, metaLAB (at) Harvard
Creative and critical engagement with AI in education. A collection of assignments and materials inspired by the humanities, for educators curious about how AI affects their students and their syllabi

AI Guide
Focused on the essentials and written to be accessible to a newcomer, this interactive guide will give you the background you need to feel more confident with engaging conversations about AI in your classroom.


From #47 of SAIL: Sensemaking AI Learning — by George Siemens

Excerpt (emphasis DSC):

Welcome to Sensemaking, AI, and Learning (SAIL), a regular look at how AI is impacting education and learning.

Over the last year, after dozens of conferences, many webinars, panels, workshops, and many (many) conversations with colleagues, it’s starting to feel like higher education, as a system, is in an AI groundhog’s day loop. I haven’t heard anything novel generated by universities. We have a chatbot! Soon it will be a tutor! We have a generative AI faculty council! Here’s our list of links to sites that also have lists! We need AI literacy! My mantra over the last while has been that higher education leadership is failing us on AI in a more dramatic way than it failed us on digitization and online learning. What will your universities be buying from AI vendors in five years because they failed to develop a strategic vision and capabilities today?


AI + the Education System — from drphilippahardman.substack.com Dr. Philippa Hardman
The key to relevance, value & excellence?


The magic school of the future is one that helps students learn to work together and care for each other — from stefanbauschard.substack.com by Stefan Bauschard
AI is going to alter economic and professional structures. Will we alter the educational structures?

(e) What is really required is a significant re-organization of schooling and curriculum. At a meta-level, the school system is focused on developing the type of intelligence I opened with, and the economic value of that is going to rapidly decline.

(f). This is all going to happen very quickly (faster than any previous change in history), and many people aren’t paying attention.  AI is already here.


 

Where a developing, new kind of learning ecosystem is likely headed [Christian]

From DSC:
As I’ve long stated on the Learning from the Living [Class]Room vision, we are heading toward a new AI-empowered learning platform — where humans play a critically important role in making this new learning ecosystem work.

Along these lines, I ran into this site out on X/Twitter. We’ll see how this unfolds, but it will be an interesting space to watch.

Project Chiron's vision: Our vision for education Every child will soon have a super-intelligent AI teacher by their side. We want to make sure they instill a love of learning in children.


From DSC:
This future learning platform will also focus on developing skills and competencies. Along those lines, see:

Scale for Skills-First — from the-job.beehiiv.com by Paul Fain
An ed-tech giant’s ambitious moves into digital credentialing and learner records.

A Digital Canvas for Skills
Instructure was a player in the skills and credentials space before its recent acquisition of Parchment, a digital transcript company. But that $800M move made many observers wonder if Instructure can develop digital records of skills that learners, colleges, and employers might actually use broadly.

Ultimately, he says, the CLR approach will allow students to bring these various learning types into a coherent format for employers.

Instructure seeks a leadership role in working with other organizations to establish common standards for credentials and learner records, to help create consistency. The company collaborates closely with 1EdTech. And last month it helped launch the 1EdTech TrustEd Microcredential Coalition, which aims to increase quality and trust in digital credentials.

Paul also links to 1EDTECH’s page regarding the Comprehensive Learning Record

 

“We need more high-impact learning practices in prison” — from college-inside.beehiiv.com by Charlotte West
Internships, apprenticeships, and work learning opportunities allow incarcerated students to keep learning after they graduate.

Maine and other states like Colorado are trying to tackle this issue through internships and employment opportunities that allow incarcerated students and graduates to put their professional knowledge and skills into practice — and in some cases, earn a living wage while doing so.

Employment and professional training opportunities inside were a major theme at the 2023 National Conference for Higher Education in Prison, where 800 educators, administrators, students and alumni from dozens of prison education programs gathered in Atlanta, Georgia last week.

 

The new apprenticeships — from jordanfurlong.substack.com by Jordan Furlong
Several American states are rewriting the rules of lawyer licensure and bringing the US into line with a key element of lawyer formation worldwide: supervised practice.

Change comes so gradually and fitfully to the legal sector that when something truly revolutionary happens — an actual turning point with an identifiable real-world impact — we have to mark the occasion. One such revolution broke out in the United States last week, opening up fantastic new possibilities for Americans who want to become lawyers.

The Oregon Supreme Court approved a new licensure program that does not require passage of a traditional written bar exam. After graduating from law school, aspiring Oregon lawyers can complete 675 hours of paid legal work under the supervision of an experienced attorney, assembling a portfolio of legal work to be assessed by bar admission officials. Candidates must submit eight samples of legal writing, take the lead in at least two initial client interviews or client counseling sessions, and oversee two negotiations, among other requirements.

Jordan mentions what’s going on in several other states including:

  • Utah
  • Washington
  • Minnesota
  • Nevada
  • California
  • Massachusetts
  • South Dakota

From DSC:
The Bar Exam doesn’t have a good reputation for actually helping get someone ready to practice law. So this is huge news indeed! The U.S. needs more people/specialists at the legal table moving forward. The items Jordan relays in this posting are a huge step forward in making that a reality.


For other innovations within the legal realm, see:

LawSchoolAi — from youtube.com

Picture this: A world where anyone can unlock the doors to legal expertise, no matter their background or resources. Introducing Law School AI – the game-changing platform turning this vision into reality. Our mission? To make legal education accessible, affordable, and tailored to every learner’s unique style, by leveraging the power of artificial intelligence.

As a trailblazing edtech company, Law School AI fuses cutting-edge AI technology with modern pedagogical techniques to craft a personalized, immersive, and transformative learning experience. Our platform shatters boundaries, opening up equal opportunities for individuals from all walks of life to master the intricacies of law.

Embrace a new era of legal education with Law School AI, where the age-old law school experience is reimagined as a thrilling, engaging, and interactive odyssey. Welcome to the future of legal learning.

 

 

 

The Beatles’ final song is now streaming thanks to AI — from theverge.com by Chris Welch
Machine learning helped Paul McCartney and Ringo Starr turn an old John Lennon demo into what’s likely the band’s last collaborative effort.


Scientists excited by AI tool that grades severity of rare cancer — from bbc.com by Fergus Walsh

Artificial intelligence is nearly twice as good at grading the aggressiveness of a rare form of cancer from scans as the current method, a study suggests.

By recognising details invisible to the naked eye, AI was 82% accurate, compared with 44% for lab analysis.

Researchers from the Royal Marsden Hospital and Institute of Cancer Research say it could improve treatment and benefit thousands every year.

They are also excited by its potential for spotting other cancers early.


Microsoft unveils ‘LeMa’: A revolutionary AI learning method mirroring human problem solving — from venturebeat.com by Michael Nuñez

Researchers from Microsoft Research Asia, Peking University, and Xi’an Jiaotong University have developed a new technique to improve large language models’ (LLMs) ability to solve math problems by having them learn from their mistakes, akin to how humans learn.

The researchers have revealed a pioneering strategy, Learning from Mistakes (LeMa), which trains AI to correct its own mistakes, leading to enhanced reasoning abilities, according to a research paper published this week.

Also from Michael Nuñez at venturebeat.com, see:


GPTs for all, AzeemBot; conspiracy theorist AI; big tech vs. academia; reviving organs ++448 — from exponentialviewco by Azeem Azhar and Chantal Smith


Personalized A.I. Agents Are Here. Is the World Ready for Them? — from ytimes.com by Kevin Roose (behind a paywall)

You could think of the recent history of A.I. chatbots as having two distinct phases.

The first, which kicked off last year with the release of ChatGPT and continues to this day, consists mainly of chatbots capable of talking about things. Greek mythology, vegan recipes, Python scripts — you name the topic and ChatGPT and its ilk can generate some convincing (if occasionally generic or inaccurate) text about it.

That ability is impressive, and frequently useful, but it is really just a prelude to the second phase: artificial intelligence that can actually do things. Very soon, tech companies tell us, A.I. “agents” will be able to send emails and schedule meetings for us, book restaurant reservations and plane tickets, and handle complex tasks like “negotiate a raise with my boss” or “buy Christmas presents for all my family members.”


From DSC:
Very cool!


Nvidia Stock Jumps After Unveiling of Next Major AI Chip. It’s Bad News for Rivals. — from barrons.com

On Monday, Nvidia (ticker: NVDA) announced its new H200 Tensor Core GPU. The chip incorporates 141 gigabytes of memory and offers up to 60% to 90% performance improvements versus its current H100 model when used for inference, or generating answers from popular AI models.

From DSC:
The exponential curve seems to be continuing — 60% to 90% performance improvements is a huge boost in performance.

Also relevant/see:


The 5 Best GPTs for Work — from the AI Exchange

Custom GPTs are exploding, and we wanted to highlight our top 5 that we’ve seen so far:

 

From the military to the workforce: How to leverage veterans’ skills — from mckinsey.com
Traditional ways of hiring make it harder for many service members to land civilian jobs. A new approach could help veterans transition to the workforce—and add $15 billion to the US economy.

This is where military veterans can make a difference. Veterans represent a source of labor potential that is untapped relative to the breadth of experience and depth of skills that they acquire and develop during their service. Members of the military receive technical training, operate under pressure in austere environments, and develop strong interpersonal skills throughout their service, making them well qualified for numerous civilian occupations. While not every military role is directly transferrable to a civilian job, most skills are—including those that correspond to US industries experiencing labor shortages, such as infrastructure and manufacturing.

 

Building a Consumer-Centered Legal Market: Takeaways from IAALS’ Convening on Regulatory Reform — from iaals.du.edu by Jessica Bednarz

Two initial high-level takeaways from the convening include:

  • Every state effort is different for a reason. There are many factors that leaders must consider when determining which model(s) of regulatory reform to pursue and how. Some examples of factors to consider include whether the state is a mandatory or voluntary bar state, whether the supreme court justices in the state are appointed or elected, and whether the model(s) to be pursued is market-based or free to consumers. The answers to these questions will likely dictate which stakeholders will likely be allies, which stakeholders will likely be opposed, and which strategy and approach to pursue.
    .
  • We need to better engage the public in regulatory reform efforts. While some organizations and efforts have included public engagement from the outset, (shout-outs to Innovation for Justice and Arizona more generally, as well as Alaska Legal Services Corporation and Frontline Justice), most have not, and this has likely hindered efforts to some degree. The data we have thus far indicates the public is interested in obtaining legal services from alternative service providers such as allied legal professionals and community justice workers, as well as through alternative business models. The more we can engage the public in this conversation, the better chance we have of creating regulatory reform that is aligned with their needs and the better chance leaders have of getting their respective efforts over the finish line.
    • Because this is an area within the regulatory reform space that is ripe for further development and impact, IAALS will soon be launching a project on consumer engagement. Stay tuned for more updates!

IAALS will share a more comprehensive list of lessons learned and recommendations for building and sustaining regulatory reform in its post-convening report, currently set to be released in early 2024.

 

Universities Can’t Accommodate All the Computer Science Majors — from insidehighered.com by Johanna Alonso
High interest in the field has led to overcrowded classes and other issues. Now some institutions are adding requirements to help force students out of the major.

Before this year, if you wanted to major in computer science at the University of Michigan, your only barrier was getting accepted to the university.

But a new model requires all students who want to study computer science—whether they are incoming or already enrolled—to apply for the major separately.

Michael Wellman, Michigan’s chair of computer science and engineering, said that the university has worked for years to try to accommodate everyone who wants to study the subject, hiring as many as six faculty members annually in recent years and even building a new computer science facility. The number of CS degrees awarded rose from 132 in 2012 to 600 in 2022.

 


When schools and families go to court over special education, everyone loses — from wfyi.org by Lee Gaines

While federal law mandates public schools provide an appropriate education to students with disabilities, it’s often up to parents to enforce it.

Schwarten did what few people have the resources to do: she hired a lawyer and requested a due process hearing. It’s like a court case. And it’s intended to resolve disputes between families and schools over special education services.

It’s also a traumatic and adversarial process for families and schools that can rack up hundreds of thousands of dollars in legal fees and destroy relationships between parents and district employees. And even when families win, children don’t always get the public education they deserve.


Future of Learning: Native American students have the least access to computer science — from The Hechinger Report by Javeria Salman

But computer science lessons like the ones at Dzantik’i Heeni Middle School are relatively rare. Despite calls from major employers and education leaders to expand K-12 computer science instruction in response to the workforce’s increasing reliance on digital technology, access to the subject remains low — particularly for Native American students.

Only 67 percent of Native American students attend a school that offers a computer science course, the lowest percentage of any demographic group, according to a new study from the nonprofit Code.org. A recent report from the Kapor Foundation and the American Indian Science and Engineering Society, or AISES, takes a deep look at why Native students’ access to computer and technology courses in K-12 is so low, and examines the consequences.


The Case for Andragogy in Educator Development — from Dialogic #341 by Tom Barrett

Understanding the Disconnect
We often find ourselves in professional development sessions that starkly contrast with the interactive and student-centred learning environments we create. We sit as passive recipients rather than active participants, receiving generic content that seldom addresses our unique experiences or teaching challenges.

This common scenario highlights a significant gap in professional development: the failure to apply the principles of adult learning, or andragogy, which acknowledges that educators, like their students, benefit from a learning process that is personalised, engaging, and relevant.

The irony is palpable — while we foster environments of inquiry and engagement in our classrooms, our learning experiences often lack these elements.

The disconnect prompts a vital question: If we are to cultivate a culture of lifelong learning among our students, shouldn’t we also embody this within our professional growth? It’s time for the professional development of educators to reflect the principles we hold dear in our teaching practices.

 

From DSC:
The following item from The Washington Post made me ask, “Do we give students any/enough training on email etiquette? On effective ways to use LinkedIn, Twitter/X, messaging, other?”


You’re emailing wrong at work. Follow this etiquette guide. — from washingtonpost.com by Danielle Abril
Get the most out of your work email and avoid being a jerk with these etiquette tips for the modern workplace

Most situations depend on the workplace culture. Still, there are some basic rules. Three email and business experts gave us tips for good email etiquette so you can avoid being the jerk at work.

  • Consider not sending an email
  • Keep it short and clear
  • Make it easy to read
  • Don’t blow up the inbox
  • …and more

From DSC:
I would add to use bolding, color, italics, etc. to highlight and help structure the email’s key points and sections.


 
 

What happens to teaching after Covid? — from chronicle.com by Beth McMurtrie

It’s an era many instructors would like to put behind them: black boxes on Zoom screens, muffled discussions behind masks, students struggling to stay engaged. But how much more challenging would teaching during the pandemic have been if colleges did not have experts on staff to help with the transition? On many campuses, teaching-center directors, instructional designers, educational technologists, and others worked alongside professors to explore learning-management systems, master video technology, and rethink what and how they teach.

A new book out this month, Higher Education Beyond Covid: New Teaching Paradigms and Promise, explores this period through the stories of campus teaching and learning centers. Their experiences reflect successes and failures, and what higher education could learn as it plans for the future.

Beth also mentioned/link to:


How to hold difficult discussions online — from chronicle.com by Beckie Supiano

As usual, our readers were full of suggestions. Kathryn Schild, the lead instructional designer in faculty development and instructional support at the University of Alaska at Anchorage, shared a guide she’s compiled on holding asynchronous discussions, which includes a section on difficult topics.

In an email, Schild also pulled out a few ideas she thought were particularly relevant to Le’s question, including:

  • Set the ground rules as a class. One way to do this is to share your draft rules in a collaborative document and ask students to annotate it and add suggestions.
  • Plan to hold fewer difficult discussions than in a face-to-face class, and work on quality over quantity. This could include multiweek discussions, where you spiral through the same issue with fresh perspectives as the class learns new approaches.
  • Start with relationship-building interactions in the first few weeks, such as introductions, low-stakes group assignments, or peer feedback, etc.
 

Innovative growers: A view from the top — from mckinsey.com by Matt Banholzer, Rebecca Doherty, Alex Morris, and Scott Schwaitzberg
McKinsey research shows that a focus on aspiration, activation, and execution can help companies out-innovate and outgrow peers.

To find out, we identified and analyzed about 650 of the largest public companies that achieved profitable growth relative to their industry between 2016 and 2021 while also excelling in the essential capabilities associated with innovation.3 Some of these companies outgrew their peers, others were more innovative than competitors, but 53 companies managed to do both. The 50-plus “innovative growers,” as we call them, are a diverse group, spread across four continents and ten industries. They include renowned brands with a trillion-dollar market capitalization as well as smaller companies that are just starting to make a name for themselves, some as young as three years old (see sidebar, “Where do innovative growers come from?”).

For all their diversity, these companies consistently excel in both growth and innovation—and they share a number of best practices that other companies can learn from.

Do innovative growers perform better than others?

In a word, yes.

From DSC:
I’m adding higher ed to the categories of this posting, as we need to establish more CULTURES of innovation out there. But this is not easy to do, as those of us who have tried to swim upstream know.

Also see:

 

New models and developer products announced at DevDay — from openai.com
GPT-4 Turbo with 128K context and lower prices, the new Assistants API, GPT-4 Turbo with Vision, DALL·E 3 API, and more.

Today, we shared dozens of new additions and improvements, and reduced pricing across many parts of our platform. These include:

  • New GPT-4 Turbo model that is more capable, cheaper and supports a 128K context window
  • New Assistants API that makes it easier for developers to build their own assistive AI apps that have goals and can call models and tools
  • New multimodal capabilities in the platform, including vision, image creation (DALL·E 3), and text-to-speech (TTS)


Introducing GPTs — from openai.com
You can now create custom versions of ChatGPT that combine instructions, extra knowledge, and any combination of skills.




OpenAI’s New Groundbreaking Update — from newsletter.thedailybite.co
Everything you need to know about OpenAI’s update, what people are building, and a prompt to skim long YouTube videos…

But among all this exciting news, the announcement of user-created “GPTs” took the cake.

That’s right, your very own personalized version of ChatGPT is coming, and it’s as groundbreaking as it sounds.

OpenAI’s groundbreaking announcement isn’t just a new feature – it’s a personal AI revolution. 

The upcoming customizable “GPTs” transform ChatGPT from a one-size-fits-all to a one-of-a-kind digital sidekick that is attuned to your life’s rhythm. 


Lore Issue #56: Biggest Week in AI This Year — from news.lore.com by Nathan Lands

First, Elon Musk announced “Grok,” a ChatGPT competitor inspired by “The Hitchhiker’s Guide to the Galaxy.” Surprisingly, in just a few months, xAI has managed to surpass the capabilities of GPT-3.5, signaling their impressive speed of execution and establishing them as a formidable long-term contender.

Then, OpenAI hosted their inaugural Dev Day, unveiling “GPT-4 Turbo,” which boasts a 128k context window, API costs slashed by threefold, text-to-speech capabilities, auto-model switching, agents, and even their version of an app store slated for launch next month.


The Day That Changed Everything — from joinsuperhuman.ai by Zain Kahn
ALSO: Everything you need to know about yesterday’s OpenAI announcements

  • OpenAI DevDay Part I: Custom ChatGPTs and the App Store of AI
  • OpenAI DevDay Part II: GPT-4 Turbo, Assistants, APIs, and more

OpenAI’s Big Reveal: Custom GPTs, GPT Store & More — from  news.theaiexchange.com
What you should know about the new announcements; how to get started with building custom GPTs


Incredible pace of OpenAI — from theaivalley.com by Barsee
PLUS: Elon’s Gork


 

 

Accenture Life Trends 2024 — from accenture.com; via Mr. Bob Raidt on LinkedIn
The visible and invisible mediators between people and their world are changing.

In brief

  • The harmony between people, tech and business is showing tensions, and society is in flux.
  • Five trends explore the decline of customer obsession, the influence of generative AI, the stagnation of creativity, the balance of tech benefits and burden, and people’s new life goals.
  • Opportunity abounds for business and brands in the coming twelve months and beyond – read Accenture Life Trends 2024 to find out more.

5 Trends

01 Where’s the love?
Necessary cuts across enterprises have shunted customer obsession down the priority list—and customers are noticing.
02 The great interface shift
Generative AI is upgrading people’s experience of the internet from transactional to personal, enabling them to feel more digitally understood and relevant than ever.
03 Meh-diocrity
Creativity was once about the audience, but has become dependent on playing the tech system. Is this what creative stagnation feels like?
04 Error 429: Human request limit reached
Technology feels like it’s happening to people rather than for them—is a shift beginning, where they regain agency over its influence on daily life?
05 Decade of deconstruction
Traditional life paths are being rerouted by new limitations, necessities and opportunities, significantly shifting demographics.

 
© 2025 | Daniel Christian