This summer, a coding class offered by a private school in Austin, Texas, was led by an unusual teacher.
The PreK-8 school, Paragon Prep, offered a series of optional, self-paced, video lessons that were automatically generated from a textbook. In them, an animated avatar made to look like the 19th-century computing pioneer Ada Lovelace taught the basics of the Python programming language.
“We’ll also look at basic concepts of data analysis, using NumPy as well as Pandas,” said the avatar in a female computer voice that sounds more like the iPhone’s Siri than like a 19th-century British mathematician, her mouth moving clumsily as she speaks. “If you have no idea what any of that means, that’s perfectly fine, good and normal. This course was meant for anyone interested in becoming a future software engineer or data scientist, not someone who is already one.”
The school’s headmaster, David McGrath, hoped the novelty of the technology would appeal to students.
“We’ve always been willing to do what it takes to get the students motivated and engaged,” he told EdSurge. “For this generation of students, an avatar is one of the tools that would work.”
This virtual version of Ada Lovelace is an example of technology known as generative AI, which is composed of algorithms that can essentially produce new content from raw information. It’s the same technology behind ChatGPT, the free tool causing alarm in schools and colleges around the country because of how easily students can use it to cheat.
But even as some educators raise concerns, others see potential for new AI technology to reduce teacher workloads or help bring teaching materials to life in new ways.
For example, McGrath sees these tools as a potential replacement for a substitute teacher—considering there’s a substitute teacher shortage in Austin, and sometimes the alternative is to turn on a movie rather than offer instruction.
“I look at it as the future of: What if we could program it to be our substitute teacher at school?” he said. “It’s almost like the teacher is programming the robot to do their job for them.”
A Booming Sector
The edtech industry is eager to build on ideas like that one. A number of startups are trying to bring so-called generative AI to classrooms as a teaching tool.
For example, the tool used by Paragon Prep came from Prof Jim, a software company that can turn existing written materials—like textbooks, Wikipedia pages or a teacher’s notes—into these animated videos at the push of a button.
“We want to make it much easier to make these teaching videos,” said Deepak Sekar, co-founder and CEO of Prof Jim, named in honor of one of his former professors at Stanford University. “Lots of surveys out there show that the latest generation prefer to learn through video, through YouTube and TikTok.”
The company hopes to work with textbook companies that will use the software to quickly create optional video versions, taught by avatars meant to embody some historical figure or modern person relevant to the material.
In a demonstration, Sekar showed how a teacher could use the software to turn a Wikipedia page about, say, the Grand Canyon into a video. The software uses a set of templates to generate a video based on the material, and it also offers the chance to edit the language on the slides behind the avatar and what the avatar is saying.
“Hit render, and it automatically creates a video,” Sekar added.
He’s not the only person thinking this way. An app called Toko helps English-language learners by serving as a conversation partner. A Swedish company called Sana Labs sells a learning-management system that promises to automatically compile courses for internal use by companies.
In the startup world more broadly, in fact, generative AI is being called the next tech boom. Perhaps the most prominent evidence of that: Microsoft is reportedly considering a $10 billion investment in OpenAI, the company that makes ChatGPT.
One reason for the interest is that the GPT-3 technology, a generative language model that can produce text that seems like it was written by a human, has drastically reduced the cost of AI features, said Matthew Tower, an education industry analyst and author of the Edtech Thoughts weekly newsletter.
“This makes AI-related features accessible to almost every edtech company,” he added.
Longtime leaders in online education say they’re bracing for a marketing onslaught.
“We’re probably right at the cusp of the marketing hard sell to institutions on the benefits of this or that AI solution,” wrote Stephen Downes, a senior research officer at the Digital Technologies Research Centre in Canada, in his Online Learning newsletter this week. He pointed to a buyer’s guide for generative tools, noting that “institutions will need to have their needs and priorities clear … before buying marking machines or teaching robots or any other such thing.”
But will educators embrace a technology that so many are complaining about as an existential threat to their profession?
While the AI technology is rolling out fast, some note that it may not actually be ready for prime time yet.
“The models do have some limitations, particularly when they are asked to provide analysis,” Tower noted.
And some educators are skeptical about the idea of avatars doing any form of teaching.
“Regardless of the quality of the presentation, this tech is not going to suddenly get the TikTok generation suddenly more engaged in their schoolwork,” said Neil Selwyn, a research professor of education at Monash University in Melbourne, Australia, in an email interview this week. “There is a reason why educational video games are not as engaging as regular video games. There is a reason why AI-generated educational videos will never be as engaging as regular videos. Brenda Laurel pointed to the ‘chocolate-covered broccoli’ problem over 20 years ago … her point still stands.”
And Selwyn is also concerned about the side effects of such technologies on the teaching profession.
“This tech makes the familiar claim that it is not looking to replace the teacher—that it will free teachers up to concentrate on high-level work with individual students. We know that this rarely turns out to be the case,” Selwyn wrote. “This tech is being primarily pitched as a money-saving device—so it will be taken up by school authorities that are looking to save money. As soon as a cash-strapped administrator has decided that they’re happy to let technology drive a whole lesson, then they no longer need a highly-paid professional teacher in the room—they just need someone to trouble-shoot any glitches and keep an eye on the students.”
Back at Paragon Prep in Austin, though, the future may already be arriving.
The headmaster there said school leaders are discussing expanding their test of generative AI to make the Prof Jim classes available as a “study hall option for students when they have a free half-hour.”