Qualcomm News:

Envisioning the future of education, and the art of learning overall, is nothing new. Over 120 years ago, French artist Jean-Marc Côté suggested how learning may look in the year 2000. While many looked at his image and chuckled — the idea that books may be somehow processed and electronically uploaded directly to the minds of young people — the concept would be re-introduced in the 1999 science fiction movie "The Matrix."

But can life imitate art? Will artificial intelligence (AI) and extended reality (XR) converge in the coming years to reflect the learning environments depicted by Côté and "The Matrix"?

K-12 education represents a $6.5 trillion industry, of which only about 2% ($16 billion) is currently invested in education technology. Despite this, many educators turn to technology to unlock students potential. But even then, technology is usually used to supplement passive teaching methods (e.g., teaching from flat interactive panels, digital textbooks, online videos and testing), resulting in little to no improvement of academic outcomes due to lack of engagement. Instead, students might be better served with active teaching methods offered through AI and XR learning experiences.

Today, most XR experiences from content providers focus on games, virtual tours, digital twins and STEM-based simulations. But with the convergence of AI modalities, we are on the verge of a true transformational opportunity to finally live up to the promise of real “personalized” learning in ways that can deliver measurable academic results. The coming years will allow us to see these various generative AI applications in use today, automating them to deliver real-time learning experiences.

What is possible with generative AI in classrooms today?

Teachers can leverage generative AI to create highly customized lessons. In a series of manual steps, educators can quickly create dynamic videos to serve as instructional tools with little-to-no technical skills. Let’s outline one such path in which such content can be created:

• Write base script: Ask Bard or ChatGPT to write you a short three- to five-minute video script about the instructional topic or challenge you wish to address. Adjust any content to address measurable learning goals.
• Transform to conversational script: Paste the initial text into a tool like Quillbot to paraphrase the content created initially by AI so it comes across as more genuine and conversational in tone.
• Create your audio narration: Paste the new conversational text into a tool such as Speechify. With their premium subscription, you’ll have a host of voices including Snoop Dogg and Gwenyth Paltrow to choose from.
• Create your video: Upload your script and audio file, with tools such as Movio or Invideo, and pick the genre and delivery format (computer or smartphone).

Using this method and many other avenues like it, educators are empowered to take their custom lessons a step further with personalization. For example, generative AI can create lesson plans that amplify a student’s strengths, while also addressing areas that need improvement. Large language models can be trained to have conversations with students for specific, on-demand support, which in turn can free up time for teachers.

But while modern technology has certainly brought education into the present day, we mostly see this in the form of tablet and computer screens. While current 2D technology takes on a form that is familiar with students today, they still end up looking at screens instead of interacting with their instructors and fellow classmates. This is where XR comes into play and seeks to address this disconnect. Another possibility with generative AI in education is combining it with the power of spatial computing to create an immersive learning environment that engages learners with interactive and collaborative lessons.



Applying the AI approach to XR learning

Today, most XR learning experiences exist as simulations or walk-throughs, an elevation of typical videos and manipulatives used in labs and classrooms. They are mostly individual experiences with a clear beginning and completion. And even with XR learning as it exists today, a PwC report found that learners mastered content four times faster than traditional classroom learning, and users were nearly three times more confident to apply the skills learned from their XR training.

What if we applied the best of generative AI into a real-time platform? The result, I believe, is a transformational tool that allows every learner to achieve their greatest potential. Let’s look at just some of the ways this concept of (Ai)daptive XR (pronounced Ay-daptive) can deliver meaningful results:

• Introduces learning where the student is currently at: (Ai)daptive XR works with the reality of that student’s ability. If they are more comfortable learning in an alternate language, then content is immediately delivered in that language (overcoming a typical obstacle for ELL learners today).
• Personalization: While today’s efforts at personalization are often no more than differentiated learning, (Ai)daptive XR can deliver true personalization. In a matter of moments, from typically the first few challenges in a simulation, the tool can recognize strength and weaknesses and adjust the learning objectives. For example, if the simulation’s goal is to measure volume of spaces, yet the learner struggles with complex multiplication, the learner would typically fail the simulation with little progress. (Ai)daptive XR can quickly assess the root cause of the gap — and alter the learning objectives to help the child meet the necessary prerequisites to accomplishing the original learning goal. A new XR simulation is created based on the learning challenges identified. On a subsequent simulation, the learner can then be better prepared to successfully complete the objectives, rather than allow these deficiencies to gain momentum and cause the child to fail academically later.
• Making learning relevant: A reason most classroom topics fail to “engage” students is that they fail to see the relevance of the subject matter. However, by understanding the personal passion of a student, (Ai)daptive XR content can be quickly adjusted to present the subject matter in a theme that is both relevant and effective to the student. For example, a simulation measuring area of farmland could quickly be updated in real-time to measuring area of fabric for a young person who daydreams of being a fashion designer. That quick change can have significant changes to the outcome of that simulation.
• Responsive to user: Rather than merely assessing performance by where one “clicks” with a hand controller, (Ai)daptive XR can respond to vocal feedback from the user(s) in their native language. Whereas today’s XR simulations have limited ability to collect user feedback, by accepting conversational speech from the user will allow for more natural interaction and help the system appropriately address questions or concerns.
• Real-time feedback: Not only do simulations provide the learner to see the actual implications for their choices, but it provides an opportunity to reinforce mastery of earlier topics. In traditional classroom settings, teachers often lack ample time to conduct assessments for their students. But with XR, real-time natural assessments can be conducted multiple times daily in a stress-free environment.
• Collaboration: Few XR learning simulations allow for multiple students to work together on a challenge, as it is easier to create individual-based content. (Ai)daptive XR can assess challenges and new scenarios introduced by the collaborative group of students, changing the content delivered in response to their collective actions in the simulation.

The possibilities are endless for (Ai)daptive XR

The evolution of instructional XR with AI will be a wonderful and welcome addition to EdTech in the coming decade. It represents an opportunity to deliver true personalized instruction for every child in every community. Such changes will require thoughtful discussions across many areas including governance, privacy, connectivity and equity.

As we continue to see examples of generative AI make its way into mainstream adoption over this next decade, the infusion of AI into our daily lives will become commonplace. Qualcomm Technologies stands at the crossroads of this sector, given its leadership in both AI and XR processing. It is my dream to see this evolve from concept to reality during my lifetime.

 Source:

The Next Frontier in Education: How generative AI and XR will evolve the world of learning in the next decade | Qualcomm

Posted By: Brink
16 May 2024