If OpenAI proves successful, the next online course you take could include a chatbot component.
Speaking on Monday at a fireside hosted by Coeus Collective, Siya Raj Purohit, a member of OpenAI’s education go-to-market team, said OpenAI may explore ways to enable e-learning instructors to create custom “GPTs” that will be linked to online learning programs .
“I hope that professors will create custom GPTs for the audience and allow people to engage with the content throughout their lives,” Purohit said. “It’s not part of the ongoing work we’re doing, but it’s certainly on the roadmap.”
Purohit says she has already observed professors uploading “semester” content to create custom GPTs using existing OpenAI tools and then sharing those GPTs with their students. “Students deal with this finite knowledge… [which] I think it’s a really effective and good way to enable them to conduct research,” she added.
OpenAI is aggressively targeting the education market, which it sees as a key growth area.
In September, the company hired Leah Belsky, former Coursera chief revenue officer, as its first chief education officer and tasked her with bringing OpenAI products to more schools. This spring, OpenAI launched ChatGPT Edu, a version of ChatGPT built for universities.
According to According to Allied Market Research, artificial intelligence in the education market could be worth $88.2 billion over the next decade. But economic growth is off to a leisurely start, thanks in gigantic part to skeptical educators.
The GPT described by Purohit may look similar to Khanmigo, the chatbot of Khan Academy, an e-learning platform launched in partnership with OpenAI last year. Khanmigo can provide students with tips on homework, test preparation, and more by integrating closely with Khan Academy’s library of educational content.
In illustrating the pitfalls of current artificial intelligence, Khanmingo makes mistakes. When the Wall Street Journal tested the chatbot in February, it turned out that he fought with basic mathematics and often did not correct mistakes when asked to re-check solutions.
Purohit, however, assured that the technology is being improved.
“All of our models are getting better and better, and our goal is to help translate them into what works for learning and teaching,” she said.
Teachers remain largely skeptical. In questionnaire this year by the Pew Research Center, a quarter of public elementary and middle school teachers said that using AI tools in education does more harm than good. Separate vote by the Rand Corporation and the Center on Reinventing Public Education found that only 18% of K-12 teachers operate artificial intelligence in their classrooms.