sourcegraph
February 8, 2023

This article is part of upstarta series of articles about companies using new science and technology to solve their industry challenges.

Since Socrates taught Plato and Plato taught Aristotle, mankind has known that the best education is provided one-on-one by an experienced educator. But this is expensive, labor-intensive and difficult to scale. The result is the imperfect classroom instruction we accept today: large class sizes, overworked and overworked teachers, and inadequate resources. Educators focus what little of their personal attention time is on the best and brightest, or on those at the bottom of the class. The wide middle ground is often left to fend for itself.

Educators may have a new tool, AI, to address these issues. The innovative form of the technology, based on computer code that mimics the network of neurons in the human brain, could reveal patterns in student performance and could help teachers adjust their strategies accordingly. The promise of an “AI tutor” — a software system through which students interact online — to give each student individualized attention could reshape education as we know it.

Among the few companies leading this transformation, reed (pronounced “rid”), a startup founded by Koreans YJ Zhang, graduated from the Haas School of Business at the University of California, Berkeley. Riiid already has a presence in the Asian test prep app market Test of English for International Communicationor TOEIC, which measures business English proficiency.Now, Riiid is entering college entrance examination with Behavior US ready market.

“Education is a complex field closely related to cognition, motivation, peer interaction, etc.,” Zhang wrote in an email. “We draw insights from learning science, cognitive biology, data science, and other related fields of research in a challenging and time-consuming iterative experimental process — which is why there are only a few players in the market.”

The first computerized tutoring systems appeared in the 1960s, presenting material as short videos that asked students questions and provided immediate feedback on the answers. Because these systems are expensive and computers are far from ubiquitous, research institutions are the main beneficiaries.

By the 1970s and 1980s, systems began to use rule-based artificial intelligence with cognitive theory. These methods guide students through each step of the problem, giving hints from the expert knowledge base. But rule-based systems failed because they were not scalable—and programming to extensive domain expertise was expensive and tedious.

When a friend of Mr. Jang evaluated such a system at Berkeley, Lee Kang Wook, now a professor at the University of Wisconsin-Madison, introduced him to deep learning, a more powerful form of artificial intelligence in which algorithms learn on their own from vast amounts of data. Mr. Jang sees deep learning being applied to teaching, where the system can learn content and student behavior over time.

He returned to South Korea and founded Riiid in 2014, working with a team of data scientists to develop an AI algorithm that tracks student performance, predicts grades, and predicts when students will lose interest and drop out.The company has released document Discuss this work at some of the world’s leading machine learning conferences.

To validate its technology and gather the data needed to improve its algorithms, Riiid has launched a TOEIC test prep app called Santa Claus (Of course, Santa collects data on children all over the world). It quickly became one of the best-selling educational apps in Japan and South Korea.

Through the app, Riiid has amassed data on student interactions, building one of the largest public education datasets in the world today, called Education Network. But Riiid has struggled to collect enough data to roll out its AI system to the wider educational arena.

“It is difficult to collect AI-trainable multimodal data in different learning environments,” Zhang wrote.

For now, the company is focused on the $300 billion test-prep market, where data is easier to collect, and is partnering with education companies around the world to develop test-prep apps. Earlier this year, Riiid partnered with Casa Grande to launch an app called OE Saber to help Colombian students prepare for the country’s Saber 11 exam.

Riiid’s success has attracted a $175 million investment from venture capital giant SoftBank Vision Fund II, bringing the company’s funding to about $250 million.

Now, Riiid is launching an AI-powered test prep platform for the SAT and ACT college entrance exams. The product, R.test, to be released in January (price not yet announced), can predict standardized test scores in a quarter of the time it takes to complete a full practice exam. By answering 30 questions, students analyze their weaknesses and receive guidance on how to improve, including relevant practice questions curated by AI. The aim, Riiid said, is for students to use the app to practice and have some confidence in their final grades when they take the real exam.

“I really like this because we can use it at home instead of having a tutor,” said Esther Yi, a parent in Georgia who tried out an early version of the platform. She found the R.test analysis to be particularly powerful. “My 10th graders will definitely benefit from it,” she said.

Oscar Torres, a Chicago high school math teacher who has tried the Riiid system, said he likes R.test because it assesses students’ knowledge in real time, rather than relying on past test scores. “As AI develops, I think it’s going to be a better and more powerful resource for us,” he said. “As teachers, we have to solve problems in real time, and artificial intelligence can help us tremendously.”

But the company’s goals are still broader than test prep. Mr. Jang said R.test is part of an effort to collect data and demonstrate the effectiveness of its algorithms in other fields. Riiid researchers continue to develop novel architectures and higher performing AI models that can track student behavior, track student knowledge, and select the best content to learn at any given time.

“Our algorithm can predict a student’s test score with astonishing accuracy, which is a number that moves like a carrot,” he said. “The more students followed the algorithm’s advice, the higher their predicted scores were.”

Zhang believes that soon teaching will no longer be based on guesswork or intuition, but on data. That may be the company’s biggest challenge: Gathering that data is a bottleneck, he added, because privacy concerns complicate data collection for schools. (Riiid says its app doesn’t collect any personally identifiable information from users.)

To address these issues, Riiid helped create the Education Safety Artificial Intelligence Alliancea global, intersectoral coalition of companies, nonprofits, and educational technology associations to develop benchmarks and standards to ensure the safe and responsible use of artificial intelligence in education.

“The dream,” says Zhang, “is to integrate these algorithms into a comprehensive system that can teach any subject to anyone, anywhere.”



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *