Imagine a world where it’s possible to build whatever online learning application you want, easily and quickly.
Say you’re head of HR at a large company entering a new line of business. How will you swiftly pinpoint all the new skills and roles required, identify the existing employees best suited to fill those roles, and design efficient learning programs customized for each employee using the best available resources?
Or you are a secondary school history teacher. If only you could quickly, easily locate the Internet’s best resources on the English Civil War, deliver them to your students along with concise summaries, create a test (including essay questions), and grade it!
AI already delivers versions of many single capabilities that could make up such applications – and, in the near future, more will become available. But there is no easy, quick, effective way to link and deploy such capabilities to build the customized learning applications and workflows we need.
One of the key insights that has emerged since we founded LearnerShape in 2018 is that the reason education is complicated – really complicated – is because it is as diverse as the people who learn and teach.
Learning tends to be done well when directed by a charismatic educator with subject expertise and direct experience of a specific learning situation, including individual student needs. It is difficult to enable such ideal learning environments online, particularly when faced with an ever-growing smorgasbord of learning resources ranging from excellent to awful with a great deal of mediocre in between.
But solving that difficulty holds huge, world-changing promise – improving learning life-long for individuals, freeing teachers to focus on key educational tasks, and delivering real applications like those imagined above. LearnerShape aims to do so with AI-based learning infrastructure.
Let me clarify what we mean by ‘learning infrastructure’. We believe this is a revolutionary concept which, like all new ideas, can be unfamiliar at first.
To take a simple analogy, let’s say you want to build a child’s treehouse. Assuming you have a child and access to a suitable tree, you’ll need at minimum some wood, nails and a tool to bang them together. These basic components are what we are calling ‘infrastructure’. But – crucially – the details of the design are up to you, including because no two trees, or children, are exactly alike.
An example from the digital world is Amazon Web Services. The huge variety of cloud computing services that AWS provides do not deliver end user applications (finished treehouses) – they deliver infrastructure (hammers, nails, planks, trees) with which users can build such applications. Although this seems like an obvious approach now, it is in fact a recent innovation that did not exist at the time of the dotcom boom and continued to be questioned less than ten years ago. In 2003, Amazon realized that there was a high-margin business in solving for other companies the same infrastructure problems that they had solved for their low-margin product distribution business. This insight has helped made Amazon one of the world’s most influential and profitable digital businesses.
The infrastructure business model is becoming increasingly common. Another good example is the Twilio Cloud Communications Platform, which delivers “building blocks to add messaging, voice and video in your web and mobile applications”. Like AWS, this is a highly profitable business model. Founded in 2008, Twilio is now worth over $60 billion.
We believe that the same approach can revolutionize learning, and we intend to be the world leader in delivering it with AI-based open source software.
LearnerShape’s mission is to make effective learning innovation widely accessible by using AI to build the world’s best open source learning infrastructure.
We also intend to generate revenue with enterprise features and by supporting the use of our open source tools. We are working with customers on a variety of such applications, including helping a large content provider enable intelligent access to their content and helping a university allow students to better engage with their course offerings.
Although various emerging technologies and platforms already try to apply computational and AI techniques towards similar ends, they deliver a mixed user experience. Owing to the complexity of learning tasks and the desire to serve diverse customer requirements, learning platforms tend to be both complex and one-size-fits-all.
LearnerShape believes that AI makes possible much more effective learning applications. Further, the power to do so should be put directly into the hands of companies, learning institutions and individuals, by providing them with open source learning infrastructure – a growing set of effective AI tools with which to create customized learning applications.
We are making our core infrastructure open source because learning is for everyone: this is a project requiring broad participation. We released the first open source version of our lsgraph library last November, providing some basic tools for enterprise learning infrastructure, and we continue to build.
I look forward to discussing what we can do together. Contact me: firstname.lastname@example.org.