Algorithms, Machine Learning, and Accessible Higher Education

Image from
omgnational.com/blog/google-the-algorithm-signals

In considering all the potential advantages and pitfalls facing us when exploring algorithmic technologies, nothing is more intriguing to me than the idea that it might unlock a more customized, accessible, and equitable educational system for the public, particularly when it comes to preschool and higher education.  As it stands, preschool and higher education are increasingly necessary to succeed academically, as children who participate in formal preschool are ahead of their peers in early grade school and young adults with a college degree have greater earning potential than those without.  But both of these educational areas are typically accessible only to those with the financial means to pay for them.  In the case of higher education, this usually means asking our nation's youth to finance their educations through loans they will end up carrying throughout a substantial portion of their adult lives.  My hope is that through innovation in algorithms and machine learning, we will be able to bring early and higher education to a wider audience for less money, and deliver a more tailored experience at the same time. 

I have a real interest in higher education administration and making higher education attainable for as many people as possible.  I also think education for education's sake should be accessible for more people, and not just education as a stepping stone to a career.  I'd love to find ways to make education possible for all those who wanted access for any reason.  I believe technology, and in particular AI, machine learning, and algorithmic technologies, have the potential to unlock that.

Algorithms are essentially a set of rules or directions that are meant to be followed.  In machines or computers, these can be incredibly complex, allowing computers to perform very detailed or complicated computations or tasks.  Computers can also use algorithms to take information they have 'learned' from performing tasks and incorporate that information into new algorithms to better improve their performance.  Algorithms are not perfect.  As human-invented constructs they are subject to the same kind of flaws, blind spots, and prejudices that human being are, so careful planning in programming must be accounted for.  But they can be an incredibly useful tool for developing ever increasingly complex machine processes that will be the key to unlocking the kinds of technologies that can be used to deliver customized educational content to students across the world for a fraction of the cost that an education carries today. 

Machine learning through building algorithms are explained simply in this short video:

Image from tech.ed.gov/highered/

By bringing education directly to the student, smart curriculum can form an important component of the educational experience and make it possible for educational information and experience to reach more people.  Machine-delivered learning will never replace human-delivered learning.  We must all absorb information from professionals in the field, learn to work and synthesize information from our classmates, and there is nothing that will be able to replace the collaborative environment of the classroom.  But for delivering content to more people, for reducing overall cost, and tailoring curriculum to the specific needs of the individual, I think that algorithm driven models of education have much to offer. 

Jordan Hutchins has an interesting take on the benefits of tailored curriculum and the implications for students in this quick video:

I am really eager to see where algorithms and machine learning take educational access over the next decade.  I hope to see an increase in educational availability, in content tailoring, in equity, and hopefully in lowering of overall educational costs to help more people afford the costs of education.