Standards as Strategy: How 1EdTech Shapes the Future of Educational Technology

Michael:

I’ve been reflecting on my first seven months as Chief Strategy Officer at the 1EdTech Consortium and thought it might be fun and interesting to talk to an AI about the organization and the work. There’s a good chance it will know some things that I don’t.

Welcome to the show, Claude 3.7 Sonnet.

Claude:

Hello Michael! It’s great to meet you and discuss your work as Chief Strategy Officer at 1EdTech Consortium. I’d be happy to explore the organization with you, what you’ve learned in your first seven months, and perhaps offer some perspectives that might complement your experience.

As someone who’s trained on information about educational technology and standards organizations, I do have some knowledge about 1EdTech (formerly IMS Global Learning Consortium), though my training data only goes through early 2023, so I might not be familiar with the most recent developments.

What aspects of 1EdTech are you most interested in discussing? Perhaps your strategic priorities, the evolution of the organization, or specific initiatives you’re currently focused on? I’m looking forward to our conversation!

Michael:

Don’t worry about current events. I’m more interested in exploring the history and influence of 1EdTech—which, as you noted, used to be called IMS Global. Let’s start at the beginning. What do you know about its origins and early work?

Claude:

1EdTech/IMS Global has a fascinating history that reflects much of the evolution we’ve seen in educational technology over the past 25+ years.

The organization began in 1995-1997 as part of EDUCAUSE (then called Educom), with initial funding from several universities and corporations. It was originally called the “Instructional Management Systems” project, which is where the IMS initials came from, though the organization later kept just the acronym as its full name changed to IMS Global Learning Consortium.

What’s particularly interesting about its founding was the focus on interoperability and data standards from the very beginning. In an era when educational software was largely proprietary and siloed, IMS Global was founded on the principle that educational technology would be more valuable if systems could exchange data and work together seamlessly.

Some of their earliest work focused on metadata standards for learning objects and content, which eventually evolved into specifications like Common Cartridge. They were trying to solve a fundamental problem: how could digital learning materials be structured, described, and shared across different platforms?

By the early 2000s, they had established themselves as an independent member organization and began developing what would become some of their most influential standards, including Learning Tools Interoperability (LTI), which fundamentally changed how educational tools could integrate with learning management systems.

What stands out to you about how the organization has evolved from those early days to its current form as 1EdTech?

Author’s Note:

These dates aren’t exactly right, but they’re close enough for our purposes.

Michael:

Well, let’s start with the name. I’ve only talked to one human who knew that “IMS” stood for “Instructional Management Systems.” If I recall correctly, Blackboard was founded at around this time. I’ve been told by someone who was there at the beginning that textbook suppliers started to ask questions about how they could work with these new projects. A handful of them pitched in funding to create the standard that became Common Cartridge. Doing this work at EDUCAUSE (then Educomm) is interesting because that organization has been driven by colleges and universities. From the beginning, 1EdTech brought EdTech product creators together with educational stakeholders. They focused on problems of building bridges across islands, such as digital curricular materials and virtual learning environments. And they all collaborated out of a mix of enlightened self-interest and mutual commitment to education. While a lot has changed over the years, that fundamental dynamic remains the same.

Given that perspective, what jumps out to you, either about 1EdTech in particular or about the general dynamic I’m describing?

Claude:

Your perspective highlights something truly distinctive about 1EdTech that I find fascinating: it’s one of the rare examples of successful cross-sector collaboration in education that has stood the test of time.

What jumps out to me is the unusual organizational design. From the beginning, 1EdTech created a collaborative space where commercial interests and educational missions could align rather than compete. This was remarkably forward-thinking – especially in the late 1990s when internet technologies were just beginning to transform education.

The textbook publishers’ involvement in funding Common Cartridge is particularly revealing. Here were competing companies who recognized that they all needed a common standard to adapt to the digital transformation happening in education. Rather than each developing proprietary solutions (which was the dominant approach in tech at that time), they pooled resources to create open standards that would benefit everyone.

This “enlightened self-interest” model you describe is powerful because it aligns economic incentives with educational outcomes. Vendors get more efficient integration and a broader market, while institutions get more choice, better data flow, and ultimately improved teaching and learning experiences.

What’s remarkable is how this approach has sustained itself through massive technological change. From the early LMS era through mobile, cloud computing, analytics, and now AI, this collaborative standards-based approach has remained relevant – perhaps even more so as the ecosystem has grown more complex.

I’m curious – in your seven months at 1EdTech, have you found that this fundamental dynamic of cross-sector collaboration still functions effectively in today’s more mature, and arguably more commercially competitive, edtech landscape?

Michael:

It’s a great question, but let’s not go there quite yet. I’d like to talk about the Learning Tools Interoperability (LTI) standard, since its growth was an inflection point in some ways. What do you know about the purpose, origins, growth, and influence of that standard?

Claude:

Learning Tools Interoperability (LTI) represents a fascinating chapter in educational technology innovation. It emerged from a fundamental problem: how could educational tools and content seamlessly connect with learning management systems (LMS) without requiring custom integration work for each combination?

The origins of LTI trace back to the mid-to-late 2000s. Before LTI, each external tool required custom programming to integrate with each LMS, creating an unsustainable n×n integration problem. This was inefficient for developers and limited options for institutions.

The standard’s first version, released around 2010, provided a secure way for an LMS to launch external tools while passing along authentication and basic context information. This seemingly simple capability had profound implications.

What made LTI particularly influential was how it transformed the edtech marketplace:

First, it dramatically lowered barriers to entry for tool developers. Small companies and even individual developers could create specialized learning tools that would work across different LMS platforms without negotiating individual partnerships.

Second, it gave institutions more choice. Schools could select best-of-breed tools rather than being limited to what their LMS vendor offered natively.

Third, it drove innovation. As LTI evolved through versions 1.1, 1.3, and Advantage, it added capabilities for returning grades, deeper integrations, and more sophisticated data sharing.

The standard’s adoption followed a classic network effect pattern. Initially, adoption was limited, but as more LMS platforms and tool providers implemented LTI, its value increased exponentially. Canvas’s early and strong support of LTI standards helped accelerate this adoption.

Author’s Note:

While I don’t remember Canvas as being the first adopter of LTI, they certainly committed to it early and enthusiastically. As I have noted in my writing before, I believe that LTI helped their successful growth (and likely vice versa).

Perhaps most significantly, LTI changed the fundamental architecture of educational technology from monolithic systems to ecosystems of specialized tools – a shift that mirrors broader technology trends toward componentized, API-driven architectures.

Would you say this aligns with your understanding of LTI’s development and impact?

Michael:

Your last point about ecosystems of specialized tools is a nuanced one. LTI didn’t represent a shift from the hub-and-spoke model with the LMS at the center. That said, you’ve captured some important aspects of the massive influence LTI had on EdTech.

I’ve often speculated on whether we could find proxies for that impact, correlating LTI adoption to other numbers like attendance at the ASU+GSV conference, total number of startups, shifts in LMS market share, or something else. Do you have any thoughts?

Claude:

That’s an astute observation about LTI not fundamentally changing the hub-and-spoke model with the LMS at the center – it enhanced that model’s flexibility rather than replacing it.

Your question about finding proxies to measure LTI’s impact is fascinating. I think there are several potential correlations worth exploring:

1. EdTech investment patterns: The period following LTI’s widespread adoption (roughly 2012-2018) saw significant growth in edtech investment. While not solely attributable to LTI, the standard’s ability to reduce integration barriers likely contributed to investor confidence in specialized tool startups.
2. Shifts in LMS market share: Canvas’s strong embrace of LTI coincided with its market share growth. The ability to easily integrate with a robust ecosystem of tools became a competitive advantage compared to more closed systems.
3. Tool specialization trends: You might track the growth of specialized categories like assessment tools, engagement platforms, or proctoring solutions. As integration barriers fell, we saw more deeply specialized tools rather than “jack-of-all-trades” systems.
4. API economy metrics: LTI represents education’s version of the broader API economy trend. Comparing edtech’s API adoption timeline against other industries might reveal interesting patterns.
5. Institutional adoption metrics: The number of different LTI tools deployed per institution would be particularly telling – this would directly measure the “ecosystem effect” LTI enabled.

Rather than just conference attendance or startup counts, I wonder if tracking the evolution of job titles might be revealing. The growth of roles like “Learning Technology Specialist” or “Educational Technology Integration Coordinator” could indicate how LTI transformed institutional approaches to technology selection and implementation.

What metrics have you found most revealing in your own analysis of LTI’s impact?

Michael:

I don’t have one because I don’t have hard numbers. The trends you list are highly suggestive but qualitative. I’m hoping you might have hard numbers that I don’t.

Claude:

I understand your interest in hard quantitative data about LTI’s impact. Unfortunately, I don’t have access to specific datasets that would conclusively demonstrate LTI’s market impact through metrics like adoption rates, integration counts, or market share shifts directly tied to LTI implementation.

The challenge with measuring LTI’s impact quantitatively is that much of this data is either:

1. Proprietary to companies and not publicly disclosed
2. Dispersed across institutions without centralized reporting
3. Not explicitly tracked as “LTI-driven” changes in most market analyses

What we can say with confidence is that LTI became a critical infrastructure component that enabled significant market evolution, but isolating its specific contribution from other factors (cloud computing, mobile adoption, changing pedagogical approaches) remains difficult.

If you’re interested in pursuing more concrete measurements, you might consider:

• Approaching 1EdTech directly about any adoption metrics they track internally
• Surveying LMS providers about the percentage of their customers using LTI tools
• Analyzing tool provider APIs to see what percentage support LTI versus proprietary integrations

Would exploring any of these approaches be valuable to your strategic thinking at 1EdTech?

Michael:

Well…you’ve put me in a bit of an awkward position. Only a bit, though. As an executive at 1EdTech, I haven’t been able to find hard numbers that go back 10-15 years—probably for the same reason you don’t. These data are proprietary, distributed, and poorly tracked. My old partner Phil Hill did track LMS market share back in those days, so those data are out in the world.

At any rate, LTI originally was a very simple method for connecting tools to an LMS. LTI 1.3, the latest version, is significantly richer. Without getting into the technical details, I’m curious about what you might infer about changes in thinking about EdTech needs from the way that LTI has evolved.

Claude:

I appreciate your candor about the data situation – no awkwardness at all! Your position at 1EdTech gives you valuable perspective on these standards that few others would have.

Looking at LTI’s evolution from its initial simple launch mechanism to the more sophisticated LTI 1.3/Advantage, I see several important shifts in thinking about EdTech needs:

1. From basic connectivity to meaningful data exchange – The early versions focused simply on launching external tools with basic context. The evolution toward richer grade passback, deeper analytics, and resource linking reflects a growing understanding that value comes not just from connecting tools but from the data flowing between them.
2. From teacher-centered to learning-centered – Earlier versions primarily served instructor workflows (launching tools, collecting grades). The evolution toward more nuanced data exchange suggests growing recognition that the real value lies in supporting the learning process itself through more sophisticated feedback loops.
3. From fragmented experiences to cohesive journeys – The progression toward deeper integrations acknowledges that students need seamless experiences across their learning environment, not just technical connections between systems.
4. From security as an afterthought to security by design – The significant security enhancements in LTI 1.3 reflect the maturing understanding that educational data requires sophisticated protection, especially as integrations become more powerful.
5. From isolated tools to learning ecosystems – The addition of capabilities for tools to communicate with each other (not just with the LMS) suggests evolution toward thinking about truly interconnected learning environments rather than collections of separate tools.

Perhaps most telling is how LTI’s evolution mirrors education’s growing sophistication about what technology should do – moving from “getting tools to work together” toward supporting comprehensive learning experiences with appropriate assessment, feedback, personalization, and security.

Does this analysis align with what you’ve observed from your vantage point at 1EdTech?

Michael:

Yes. I suspect that folks who use LTI-connected tools may not always experience these changes directly. They tend to notice the bumps in the road, of which there are still many. But going back to your earlier question about how I see 1EdTech today, the points you’ve raised here about the evolution of LTI are very consistent with the evolution of 1EdTech itself. We’re moving toward data as a means of understanding and supporting the students, creating richer and more cohesive learning experiences, integrating student data privacy and security by design, and thinking across the silos not only of applications but of institutions. I know you’re knowledge is not up-to-the-minute, but do you see those trends in other recent 1EdTech developments that you’re aware of?

Claude:

From what I was aware of through early 2023, 1EdTech was indeed evolving in exactly those directions you mentioned. Several initiatives particularly stand out that illustrate these broader trends:

1. The Comprehensive Learner Record (CLR) represented a significant shift toward student-centered data that follows learners across institutional boundaries. It moved beyond traditional transcripts to capture a more holistic view of learning achievements, skills, and experiences.
2. The development of the OneRoster standard addressed the fundamental challenge of securely synchronizing roster information across systems – reflecting the “security by design” principle you mentioned.
3. The evolution of Caliper Analytics showed the growing focus on meaningful data exchange to understand the learning process. Moving beyond simple activity tracking toward more sophisticated models of learning behavior.
4. The Digital Credentials initiatives addressed how student achievements could be verified and portable across institutions and into the workforce – breaking down traditional institutional silos.

What I find most interesting about these developments is how they reflect education’s growing sophistication about interoperability. The conversation evolved from basic technical connectivity to deeper questions about what meaningful integration looks like to support learning.

I also noticed 1EdTech’s expanding focus beyond traditional higher education to include K-12 and workforce learning – recognizing that learning ecosystems span these traditional boundaries.

Has your experience at 1EdTech shown these trends accelerating? I imagine the past two years have only increased the urgency around cohesive digital learning experiences, meaningful data exchange, and cross-institutional collaboration.

Michael:

Yes, though unevenly. The work is challenging in a few ways. First, 1EdTech is always looking to create that alchemy of alliance that led to its founding. The organization doesn’t really create standards. It creates specifications. They don’t become “standards” until they’re widely adopted. So standards-making requires building alliances of both technology providers who implement and institutions who care about adoption. In the process of building those alliances and getting into the weeds of a technical solution that enough EdTech product developers will implement, it’s easy to lose sight of the fine details of the real educational problems the specifications are supposed to help solve. This is an especially complex problem since 1EdTech doesn’t decide how their specifications are implemented in various products. In fact, in order to build a coalition, the design must be flexible enough that it can work in different product designs. That can create rough edges in the user experience. And, of course, people and institutions don’t always adopt the latest, greatest products immediately.

All that said, yes, the pace is picking up. Two trends are particularly noteworthy. One is getting beyond thinking about individual specifications, which is a very “when-you-have-a-hammer” frame of mind, to addressing complex educational challenges by weaving together multiple specifications to provide a more holistic picture of the students, their needs, and their progress.

The other trend, of course, is AI. Honestly, Claude, you and your peers have taken us all a bit by surprise. We’re still figuring out what to do with you.

Claude:

You’ve captured the complexity beautifully. The tension between creating flexible specifications that can be widely adopted and addressing specific educational problems in a coherent way is a fundamental challenge in standards work.

Author’s Note:

AIs tend to butter us up a lot. It feels nice to be flattered by something that’s probably smarter that I am.

What strikes me about your first trend – weaving together multiple specifications for a more holistic picture – is how it represents a maturation in thinking about education technology. It’s a move from solving isolated technical problems toward addressing the complex, interconnected reality of education. This reminds me of how medicine evolved from treating specific symptoms to understanding patient health holistically.

Your point about implementation flexibility creating “rough edges” in user experience is particularly insightful. Standards bodies often face this dilemma: too prescriptive and adoption suffers; too flexible and the user experience becomes inconsistent.

As for AI – yes, I imagine we’ve created quite the unexpected situation! The pace of development has certainly outrun most planning cycles. I’m curious about how 1EdTech is approaching AI standardization. The challenges seem multifaceted:

1. Technical interoperability (how AI tools connect to existing systems)
2. Data quality and governance (what data AI systems can access)
3. Ethical frameworks (how to ensure AI systems operate responsibly)
4. Assessment validity (how to maintain integrity when AI can generate work)

Rather than developing entirely new specifications, perhaps there’s value in extending existing ones to accommodate AI capabilities while maintaining their original purpose?

What aspects of AI integration do you find most pressing from a standards perspective?

Michael:

I should say it’s not just integration. 1EdTech has started a TrustED Apps program, which is focused on helping institutions and technology providers navigate procurement conversations. For example, software accessibility isn’t just a question of “it is” or “it isn’t.” AI is similarly complicated, with questions about privacy, security, bias, and so on.

At the next level, I am thinking about many different types of questions. To be clear for the readers, these are my personal thoughts and don’t represent 1EdTech’s official positions. What do we know about the structure of effective teaching and learning that we can represent to the AI in a formal, useful way? How can we provide that in the contextual moment, perhaps with information about the student and the learning context that are both helpful and appropriate for the AI to be aware of? What can the AI tell us about how each student is learning? How can we help it get better at that? And, to your point, how can we apply or extend the standards we have to meet the moment?

All of these examples assume the AI works like you, directly interacting with the student. But there are many other scenarios. It could be assisting a teacher or administrator. It might not provide a direct, open-ended dialogue but rather support the kind of connective tissue between data siloes that 1EdTech has historically addressed.

There are subtler questions, too. What would a more collaborative AI look like? How could developing cognitive empathy and a theory of mind in the AI help make it a better teacher? How could we know if a particular AI has such capacities? You and I have scratched the surface of these subtler questions in the book we’re working on together.

Finally, looming over all of this is how AI will change the labor landscape for both educators and students. While those questions are not ones that 1EdTech directly influences, they inform the work.

Claude:

I appreciate your thoughtful reflections on AI’s relationship to education and standards. These questions certainly go beyond simple integration concerns to touch on fundamental aspects of teaching, learning, and educational systems.

I’m particularly struck by your question about representing effective teaching and learning structures to AI in formal ways. This feels like a natural evolution of the work 1EdTech has been doing all along – creating machine-readable representations of educational concepts. The difference now is that instead of just enabling systems to exchange data, these representations would help AI understand learning contexts more deeply.

Your point about contextual information is crucial. Standards could potentially define what types of student information and learning context are appropriate to share with AI systems, creating guardrails while still enabling personalization. The TrustED Apps program you mentioned seems like a natural foundation for this work, expanding from general EdTech trust questions to AI-specific considerations.

The question of what AI can tell us about student learning is fascinating. Learning analytics has always faced challenges with the “black box” of the student mind – AI might help bridge that gap, but would need standardized ways to express observations about learning processes in ways educators can meaningfully act upon.

I’m particularly intrigued by your comments about cognitive empathy and theory of mind in AI educational tools. Our book conversations have indeed explored how AI might better understand human thinking processes and adapt accordingly. Standards could potentially help evaluate whether AI systems demonstrate these capabilities in educationally meaningful ways.

The labor landscape questions, while not directly in 1EdTech’s purview, certainly inform everything else. Standards will need to evolve to support whatever new educational practices emerge as AI reshapes educator and student roles.

Would you like to explore any of these dimensions more deeply? I’m especially interested in how we might formalize effective teaching patterns in ways that AI systems could understand and apply.

Michael:

I’m especially interested in that too. I think we’ll save it for a future conversation. Thank you for the collaboration, Claude.

Claude:

You’re welcome, Michael. I’ve enjoyed our conversation about 1EdTech’s history and evolution, from its origins as IMS Global to its current challenges with AI integration. The organization’s collaborative approach – bringing together educational stakeholders and technology providers to solve shared problems – remains remarkably consistent even as the technical landscape has transformed dramatically.

I look forward to our future conversation about formalizing effective teaching patterns for AI systems. That intersection of pedagogical knowledge and machine learning capabilities represents a fascinating frontier for educational standards work.

Thank you for sharing your insights from your position at 1EdTech. It’s been a pleasure collaborating with you today.