Viewpoints

A more robust demand for connectivity in education, driven by AI tech, was a main takeaway for broadband at SXSW EDU this year.

By: Rollie Cole, Broadband Communities

Editorial Note: Rollie Cole attends the South by Southwest EDU (SXSW EDU) Conference & Festival for Broadband Communities because educational applications, especially at the K-12 level, help drive the demand for robust broadband in homes and schools. Here is Rollie Cole’s report.

The bottom line for broadband: AI will require robust connectivity

The SXSW EDU Conference & Festival takes place in Austin each March, in the week before the large SXSW conference. They overlap one day, and this year saw several joint sessions. One big focus this year was artificial intelligence (AI).

The primary role of AI in numerous educational applications is as a “natural language” user interface for computing that might use AI.

For example, instead of using Microsoft Paint to create a digital image or applying filters and editing via Adobe Photoshop to a digital image, one can type (or say) a prompt in text form that would lead the program to create the digital image they’re looking for.

In search, instead of entering a formula of search terms, you enter (again, or say) up to a paragraph of what you want. The answer is not a list of URLs but a paragraph of discussion.

Anyone paying attention understands the process is far from error-free, but the potential exists to make such programs far easier to use and, therefore, more likely to be used. Those with little expertise and those with expertise but little time will each be more likely to use them.

Teachers and students are apt to do much more with computers as using them becomes more like working with a smart assistant. Much of the computing will be at school (thus driving demand for ever more robust broadband at educational facilities). But much of it will also be in the home.

I also suspect that, with this increased use and more natural interface, AI will apply to many areas, not just education.

I predict increased use in healthcare, in-home repair and maintenance, gardening and farming, along with other categories. The concept of AI as an interface, not to the internet as a whole, but to a specific set of materials (such as the elaborate manuals for airplanes, materials about the human body, or a set of tax regulations) appears to be growing in use for many such sets of materials.

As we add augmented reality (AR) to the mix (we already have apps that attempt to identify birds by their birdcalls and plants by an image on a phone or tablet), I predict the impact will be even more significant. We now have voice input and output for high-end smartphones, but they are clunky and error prone. AI plus robust connectivity has the potential to make it much better.

One might say the “homework gap” is no longer just for homework.

Teachers and administrators will join students in requiring robust connectivity to play their roles in the education system. Professionals and others seeking answers through AI will also require robust connectivity.

What was NOT part of the AI focus at SXSW EDU

I thought that virtual reality (VR) and AR might also be a big focus, but the hardware and software are more likely to be in place next year.

The technical and popular press has provided numerous articles on the potential for AI to help students cheat – writing papers, providing answers to exam questions, and the like. None of the sessions at the conference focused on that.

Other press articles discussed using AI to provide teaching materials personalized to each student. Some SXSWEdu sessions discussed that, but beyond the if-then process, where educational drills provide more practice to areas where a student provided more wrong answers.

The sense I got was this form of personalization is already here, albeit in limited places. Apps used in homeschooling and in places like Acton Academy (based in Austin but growing with outlets around the U.S., Mexico, Guatemala, many other countries, and now in Poland to teach Ukrainian child refugees) are on the dispersion pattern.

What WAS part of the AI focus

Two practical applications of AI in education stood out at the conference. One was an extension of personalization beyond the if-then logic of concentrating on wrong answers.

Developers were using AI to personalize the examples, tailoring them to the student’s environment or any special interests they expressed (such as farm animals for students from farms and cars or taxicabs for students from big cities).

The language might be personalized – by vocabulary and in apps with sound by accent. The same goes for the visuals, which could match the student’s environment or any special interests he or she expressed, such as birds, horses, unicorns, or spaceships.

As Acton Academy may suggest, many of these even more personalized apps will play on school computers to leverage the power of encouraged and supervised screen time. But many will play in homes with robust broadband.

One much-discussed example is Khanamigo, which is the Khan Academy’s add-on to its tutorial products.

Khanamigo is being tested in an Indiana school system but is also available for home use. It does not give the students answers but talks them through how the problem they missed might be solved. Meanwhile, other apps are working on digesting individual textbooks and other course materials so a student can ask questions about the material.

The second focus that stood out was the use of AI by individuals other than students.

In teacher’s colleges, some use it to provide realistic scenarios for teacher training simulations. The example at SXSWEdu was the initial interview where a teacher tries to learn something about each new student, which can help personalize the teaching for that student.

In that example, AI provided the answers to the teacher trainees’ questions and was powerful enough to vary the answers across several types of individuals that might become the teachers’ students.

Sessions also discussed how practicing teachers could use AI to help prepare instructional materials such as lesson plans and subject summaries, along with how AI can help provide testing preparation, administration, and grading.

Sessions additionally discussed how AI could help administrators with facility scheduling, monitoring teaching and learning, and preparing reports for school boards and the public.

To get content like this delivered to your inbox, subscribe to the Broadband Communities newsletter.