top of page

Education and AI

Updated: Aug 23, 2022

The COVID pandemic has exposed and exacerbated deep social issues, including our failure to teach algebra. But what does algebra have to do with pandemics? The answer lies in the difficulties of getting people to take the virus and basic preventive measures seriously. The vaccines have been a tremendous scientific achievement, but as always, the biggest tool we have is our behavior. That behavior would be easier to influence if people knew when and how to utilize relevant math reasoning.

People don’t see the somewhat counter-intuitive FACT that small changes in behavior can have large scale consequences. In particular, many people don’t understand the amplifying effects of exponential growth. A small number of infections can rapidly lead to a large outbreak. Fast? No, STUNNINGLY fast. Many people don’t get this in a visceral way that leads to actual behavior change.

The absence of understanding the meaning of exponential growth, a topic that students typically first encounter in algebra classes, is crippling our individual and collective efforts to deal with vital, even existential issues. The cliched rip on algebra is that it’s never used in real life. People seem to think that algebra is a lot about polynomials (which most people don’t use) but it’s actually about thinking. Saying that algebra is unimportant because we don’t use polynomials is similar to saying that reading Moby Dick is pointless because most of us never work on whaling ships.

The sad performance of American education in algebra pierces awareness every so often in headlines about our performance with respect to international standards. Given algebra’s crucial role in STEM education, this has led to a monumental response (both in rhetoric, research and dollars). There have been many ideas promulgated, but technology, partly due to large investments in an emergent EdTech sector, has been an enduring part of the overall effort to improve education. Technology’s role has been controversial, with some advocates pushing disruptive opportunity, and detractors deeply skeptical.

What’s been the result? The basis for international comparisons, the PISA, is just one measure but SO FAR the results are not encouraging. US performance remains quite mediocre, and the gap between the higher and lower performing countries is widening. Of course blame has been widely distributed, technology getting its fair share, some of it quite vitriolic and intense. Despite that, technology has stayed firmly in the mix.

There are good reasons:

Scaling what works:. Even if non-technologically supported efforts work, it’s hard to imagine how to make large-scale improvements to STEM related education without technology. You can scale technology. It’s much harder to scale great teachers.

Failure is not the only story: There has been very real incremental progress in technology-based education efforts and there is actually evidence that technology can make a difference. Why this hasn’t translated to broader success is a complicated issue. Important organizations like Digital Promise are guiding and shaping the steady accumulation of progress. Technology takes time to develop and work through the details to move heaven and earth. Amara’s law (also known as Gates’s law) says that we tend to overestimate the short term effects and underestimate the long term effects of technology. Education is a complicated ecosystem driven by layers of conflicting accountability measures and limited resources. Just because progress is slow, doesn’t mean that the impact of technology won’t one day be fast (and profound).

The Pandemic Has Been a Jolt: COVID has made technology necessary and has forced Edtech companies and schooling systems alike to determine core learning and delivery mechanisms that can be effectively automated. Its destruction of in-person education has forced disciplined thinking around new teaching models and tools. Investor friends of ours have reported order of magnitude changes in demand for the services of their portfolio companies. This has been a sink or swim moment for education. Just as COVID has caused companies to rethink the role of the office, will COVID cause some rethinking of how technology can be used and how school time can be organized?

Unfortunately, technology solutions have fallen short, and progress from here will depend in part on understanding why. Technologists undoubtedly overpromised and there was some credulous buy-in that led to disastrous deployments. Cycles of wild hope followed by disillusionment are not unique to education. The effects of failure in education, though, are more destructive. A corporation can abandon a failed IT project. We cannot abandon eighth graders who were harmed by a misguided reengineering of their learning time.

Overreach certainly has been an issue, but there are substantive reasons as well:

Ecologically bad solutions: Too often technological solutions were introduced in wholesale fashion without careful design for the context (i.e. content & pedagogy, classroom feasibility, effective teacher preparation & support).

Unrealistic models of mind: All learning technology has an explicit or implicit model of the mind. Technological solutions tend to treat the mind as a computer. Responses to a program are neatly categorized whereas in reality student responses are often fuzzy, partial, and they learn through refining this rough draft thinking over time. This difference requires very different styles of interaction.

Limited imagination and the McLuhan problem: There has been lots of technology that mostly reproduced old educational models online, like publishing videos of lectures, which historically haven’t produced outcomes for the vast majority of learners. Digitizing these old forms didn’t change that fact much except for already advantaged learners.

History is hard to figure out when you are living in it but we believe that the times, they are a changin’. These changes may yet allow us to deliver on the promise. We are encouraging optimism, not techno-euphoria, careful evaluation not withering cynicism.

Among the many changes there are three in particular that make a new paradigm for the use of technology possible in education.

  1. Understanding of Human Learning: Over the last few decades we’ve seen a flowering of research on human learning. These studies now even include direct observations of brain function. We now know a lot more about learning and what works.

  2. Software Capability: Systems are now capable of learning over vast amounts of data, interacting with people through more natural interfaces and using richer models of users. We now expect our software to evolve, expanding and improving its fit and function.

  3. Assistive Approaches: We are many years post-Siri, and now assistive approaches delivered through apps are more capable. Moreover, the younger population (students and increasingly teachers) now integrates technology smoothly into their normal activities. Users are more prepared to use technology effectively.

A new paradigm for educational technology is emerging as a result of these factors. It is very congruent with the views of the Russian psychologist Lev Vygotsky who thought of students and children as active agents and the role of teachers and parents as providing support (typically called scaffolding). Scaffolding creates a ‘zone of proximal development’, what learners can do with assistance. Reflect on how parents ‘teach’ their children to walk. They don’t lecture or try to show them how to walk. They assist the learner’s internally driven efforts by, for example, holding their hands and offloading some of the balance issues that might be too hard to take on when the child is learning the basic mechanics.

Technology has become much more capable of offering authentic experiences that foster active learning, and providing integrated scaffolding to students. Rather than introducing technology as an artificial add-on to an already overburdened educational environment, high value assistive functionality can now be integrated more naturally. Companies (including our own efforts) like Labster, 3D Bear and Prisms are utilizing tactile immersive environments for students to physically interact with their environments and build core knowledge experientially. And similarly, Amira Learning and Reperio are using AI to analyze and repair performance gaps. We see these ventures and others as part of a Cambrian explosion of new models and approaches.

Technology in education has often been associated with ‘robot teaching’, replacing work done by teachers by computers. We have seen first hand how the work of teachers is often driven towards tasks and goals that are quite at odds with the deeper motivations to teach. This is mostly due to the fact that the teaching role is an entirely impossible job - she is responsible for delivering content in an engaging and pedagogically effective way, providing feedback to 30 students simultaneously, the duties of a psychometrician, counselor, and grader amongst others like lunch duty. This has led to highly destructive attrition rates.

In this new paradigm, that is not the guiding inspiration. An assistive approach may allow teachers to recover their more direct and satisfying role. Quite in contrast to the idea of replacing teachers, we think that we can follow a technology direction that makes teachers’ professional lives BETTER by giving them tools that facilitate effective learning models at scale.

New technology allows us to imagine new things, new opportunities - it’s like a new mental grammar. It does not, though, ensure the success of any of those ideas, especially in human systems like education. In such systems, the fit of the technology with human values and characteristics is utterly important. We are arguing that technological progress has reached a point where that fit can be realized. We are not looking for a silver bullet, but we believe that the rise of assistive approaches has a chance of making a meaningful contribution to an existentially important need.


Recent Posts

See All

We’re going to use this space for short takes on current issues in AI that we believe are relevant to our customers. AI is a big, complicated topic and the term has become confusing. AI is much mo

It’s now a cliche that data changes every field it touches. But how? It might be that it just makes the business better, maybe more scalable. But there are lots of cognitive and cultural effects.

As AI techniques for modeling big data (especially deep learning) have exploded in interest and use, concerns over bias and negative impact have been raised. There are many articles and many general

bottom of page