Education stands to benefit from rapid developments in artificial intelligence. But historically, adaptive learning software has been programmed in a top-down fashion. It asks a question, and if the child provides a particular answer, a set of prompts or tips might be shared or a new (perhaps easier, perhaps more advanced) question might be asked. Whether this is or is not AI is up for debate; however, the development of such programs is labor intensive and generally only effective in providing practice for concepts taught in class.
As the Innovation Leader at an international education company, I work with both educational leaders on leveraging digital learning ecosystems and with education technology companies to improve their products for use in our schools. As a classroom teacher many years ago, I used basic adaptive technology, and I've always encouraged the use of these technologies within the classroom and at home. While this has led to some increased customization and more individualized learning pathways and pacing, I still believe top-down AI has fallen short of enabling transformation.
Advancements in bottom-up AI, such as the new Google Translate, and the accessibility of neural networks suggest that much more may be possible soon, as captured by the recent New York Times article "The Great A.I. Awakening."
For schools, this may allow for a few unique advancements in particular:
- Unifying data in different silos to build school-specific predictive models.
- Providing adaptive learning programs that respond to students’ emotional reactions and/or demonstrated learning biases not just keystrokes and answers given.
- Presenting students with multidimensional performance data and personalized pathways for remediation, enrichment, or course selection.
- Providing teachers with up to the minute data on the efficacy of learning interventions based on national or even global data.
- Content that is filtered, updated and presented independently, not when textbooks are changed or websites are updated.
- Identifying bias or in textbooks and other instructional materials.
- Coaching for students based on a range of factors from performance to specific learning needs..
- Curation of data for third-party education or service providers that interact with the school.
Imagining An AI-Supported School
An AI-supported school would look quite different from the schools we attended and that our children attend today. Let’s imagine Ashley, a student at our imaginary AI-infused school. Ashley wakes up and 6 a.m. and confirms by SMS that she will be taking the bus. As she gets ready for school, she receives an SMS with her transport details including pick-up time and drive time. This may vary by up to 10 minutes each day, as the route is calculated based on confirmed attendance and traffic to minimize the amount of time spent on the bus. When she arrives at school, she meets with her tutor to review her learning plan for the day. She’s been struggling with algebra and the tutor has been given a possible treatment plan based on Ashley’s strengths and national data on the most effective academic intervention routes.
Ashley moves freely through the school over the course of the day and receives a message at 2 p.m. that the engineering lab has only four students and the teacher has arrived for office hours. She has a project due in a week, but hasn’t been in the lab in several days - a red flag that prompted the SMS. Later, during study hall, she receives a ping that her friend Alice is nearby and could use some help with Spanish, one of Ashley’s strengths. When the fire alarm goes off in the afternoon, students each receive an SMS based on their location and the best pathway to exit the building. Teachers and staff meanwhile receive instructions based on the number of students in their area at that moment.
While this AI-infused school may sound like a futuristic dream, the technology for such a school experience exists.
The Path To Effectively Using AI For Education
Unfortunately, current practices in most schools make the transformation of education through AI unlikely. Below are five actions that schools, data scientists, and ed tech entrepreneurs can take to work as partners and accelerate the rate of AI adoption and integration.
- Collect and share more data. Effective AI is reliant on data -- lots of data. While schools tend to be data-rich environments, there are large gaps in the data that could be collected. All too often schools collect just data in the big three: performance, attendance and behavior.
- Engage students in discussing AI. The 2016 WEF raised 9 ethical concerns regarding AI. Conversations about AI are a great springboard for debates, critical thinking and contemplating morality. We're trying to do this ourselves: My company and its partner recently launched a twelve-week program in which students learn about exponential technologies such as AI. Through this curriculum, students are encouraged to consider how these technologies might impact the future -- and even change the way we think of ourselves. Any school could do the same.
- Add data science to the curriculum. Not only will this serve to close the global employability skills gap, it will increase the number of people, both students and adults, within a school who are empowered to work with data. The LA Unified School District did this in partnership with UCLA 2014, and organizations like Microsoft and Coursera offer low-cost data science programs open to anyone.
- Play with AI. Encourage teachers to make a free Amy Ingram account to handle scheduling parent-teacher conferences. Have a Siri or Alexa station in the classroom for just-in-time information gathering. Try out the AI possibilities that already exist.
- Use AI to focus on teaching and to outsource administrative tasks. AI is likely to spread first to operational departments within schools, such as human resources and transportation. Engage teachers and school leaders in thinking about how this might help them. Activities such as linking data sets including student behavior, time spent in school transport, and attendance to identify potential roots of learning difficulties or even creating differentiate materials can easily be done without human intervention.
AI will not replace teachers. Computers will not replace schools. But we can all work together to use to use the newest technology to improve education.
