Teacher-Trained Artificial Intelligence to Teach Students in Brazil

Priscila Zigunovas
Adjust the
text size

Quick take: Several innovations are taking place at Saint Paul Business School. The unprecedented use of cognitive computing is just one of them. The institution also aims to democratize education in Brazil and transform executive education.

São Paulo, Brazil

Imagine what it would be like if professors at your institution had a new career possibility: to teach reasonably complex topics to an artificial intelligence capable of understanding student questions and explaining to them various concepts, techniques and relations within a subject. What kind of implications would that have on the future of education? This may resemble an episode of Black Mirror, but it’s actually happening at one of Brazil’s top business schools which intends to revolutionize executive education – and maybe adult education itself.

Saint Paul Business School, based in São Paulo and five times listed among the best in the world in the Financial Times ranking, spent all of the last year developing a platform called LIT which uses artificial intelligence in an unprecedented way. The project began to be executed in November 2016, with the contribution of approximately 40 vendors.

Onlearning: The Concept Behind the Platform

José Cláudio Securato, Saint Paul’s CEO and Founding Partner, explains that the institution has developed a concept called onlearning, which proposes a theoretical basis for how they believe the future of executive education will be, and also adult education in general – education for people in the job market. “LIT is the way we found to develop a product based on onlearning,” he explains.

But what does onlearning consist of? According to Securato, although the idea of lifelong learning is familiar to the different generations currently in the workforce, there’s something new in the air. People tend to have grown up hearing that studying is important, but learning tends to occur at certain times throughout life, not continuously.

Photo José Cláudio Securato, Saint Paul’s CEO and founding partner
José Cláudio Securato, Saint Paul’s CEO and Founding Partner

“Our generation’s life-long learning is on and off, as if we turned on and turn off our learning mindset, as if there was a moment to learn and another moment when you don’t have to. That’s what happens today. But we think that’s outdated and that the learning process should be ongoing, like we were always on ‘learning mode’. That’s where onlearning came from,” explains the CEO, who wrote a book called Onlearning: How Disruptive Education Reinvents Learning (OnLearning: Como a Educação Disruptiva Reinventa a Aprendizagem, available in Portuguese only).

For Securato, learning doesn’t have to be monotonous. On the contrary. “It has to be more organic, more fluid, more intuitive. It has to be compatible with our modern life and with ourselves. Onlearning disrupts the traditional model of learning, which is what we are proposing,” he says.

How Does LIT Work?

LIT is like a Netflix of courses. Students will pay 99 Brazilian reals per month (approximately 30 dollars) to get access to all course content available. According to the institution, even if you were willing to study eight hours a day, 365 days a year, you wouldn’t be able to consume more than 15% of the content which will become available on March 1st, the day the platform is set to be launched.

“We also have 1,500 exercises and case studies solved step by step, 7,000 books on our digital library and 10 MBA programs. It is, in fact, lifelong content,” says Securato. The platform allows students to earn micro-certifications along the way, so they can acquire specific skills they need or want to develop. Upon completing these certifications, the learner has the option of completing a set of additional components in order to earn an MBA degree from Saint Paul, recognized by the Brazilian Ministry of Education, for a fraction of the cost of a traditional program.

Although delivering course content through a subscription platform is already an impressive innovation for a business school, Saint Paul decided to take it one step further. They also wanted student experience to be personalized.

In order to do that, Saint Paul has invested heavily in technology. They have combined the artificial intelligence technology IBM Watson with e-learning resources from Blackboard’s digital learning environment, as well as videos, a library, and a social network that connects everyone to everyone – you can’t add “friends” – used by students to talk about complex topics and share their experiences and knowledge in business.


Discover the features that allow institutions to deliver a more personalized learning experience

Inside LIT, Watson is called Paul. Taught by Saint Paul professors and experts, Paul has learned a lot so far. He is able to identify the students’ strengths and weaknesses and personalize their learning. He also analyzes the students’ textual production and interactions in order to capture personality traits and suggest the most efficient ways for them to learn. And he can teach by talking to students and answering their questions.

“Paul is a supplementary method of learning, which doesn’t exclude the others. The platform contains a myriad of learning methods and Paul is one of them, a disruptive, innovative method, which it’s especially interesting because it allows learning to be unstructured, meaning that the student can start from wherever he or she wants,” says Adriano Mussa, Academic & Artificial Intelligence Research Director and Partner at Saint Paul Business School.

Paul Trainers

Founded and managed by teachers, Saint Paul prides itself of putting students first and professors right alongside them. Securato and Mussa make a point of emphasizing that Paul can’t replace a professor. In fact, Paul has created a new role for the teaching staff: the institution already has professors whose title is “Paul trainer.”

“The professor teaches Paul, Paul teaches the student, the student teaches Paul and Paul has to teach the professor, who will continuously work in that content curatorship, since content isn’t linear, but cerebral, practically a neural network. So, the complexity of content curatorship is another role for the professor. What we are seeing is that Paul actually has expanded possibilities for professors in terms of professional performance,” says Mussa.

“The professor teaches Paul, Paul teaches the student, the student teaches Paul and Paul has to teach the professor, who will continuously work in that content curatorship"
Photo Adriano Mussa, academic & artificial intelligence research director and partner at Saint Paul
Adriano Mussa, Academic & Artificial Intelligence Research Director and Partner at Saint Paul

According to Securato, if the work of the professor is going to change, Saint Paul is getting ahead and saying how it’s going to change. With Paul’s evolution, according to the CEO, professors will have time to study more, learn more, teach more and guide students in more complex topics.

In Mussa’s opinion, the use of artificial intelligence in teaching is a natural advancement for society, because if the student is able to learn less complex, or reasonably complex subjects from Paul, then the professor can focus on teaching more advanced contents.

Additionally, subjects related to the humanities remain a challenge for artificial intelligence, so the work of human professors is even more essential in those areas.

Democratizing Education

One of the greatest barriers for Brazil’s growth is the low efficiency level, as measured by any economic index, as Mussa points out. “So, if one of our major problems is inefficiency, Paul is helping us to take bigger steps as organizations, as society and as a nation. That’s what Brazil needs: high quality, personalized, democratized education. That’s our greater purpose, our role in Brazil’s social and economic change,” he says.

“When we consider all the innovative features that LIT offers for 99 reals, we know for sure that we are democratizing quality learning, which becomes accessible to millions of people that maybe wouldn’t be able to pay 40,000 reals for a Saint Paul MBA program. So it’s a profound change we are making,” says Securato.

Meet Paul, the Artificial Professor

An omnipresent professor that can talk to students any time they need, explain simple to reasonably complex concepts and relationships, disambiguate terms, and personalize learning according to each student’s personality. Paul is a very interesting idea and is one of a kind, to say the least – so far, no other educational institution has been reported to be using artificial intelligence in the same way. But how exactly does Paul work?

Development: Paul was developed based on the technology IBM Watson. To allow professors without any coding knowledge to train him, a team of Saint Paul’s experts created their own algorithm.

Training: Before Paul was ready to teach students, he needed to be taught. Experienced professors have spent 12 to 18 months learning how to do that. The first subject that Paul learned successfully was accounting. The goal of the institution, over the years, is to transfer all their knowledge over to Paul.

Cognitive Computing: Paul is a new way to learn inside a platform with multiple learning objects. The student can learn by talking to Paul and asking him questions. When a student asks Paul a question, he uses natural language processing to understand what the student wants to know, and delivers an answer prepared by the professors.

Unanswered Questions: If Paul doesn’t have the answer the student is looking for, real instructors are available to help. However, when a question is unanswered, a process that Securato calls “content curatorship” begins to occur: a team of professors works quickly and continuously to teach Paul the new content, so the next time that question is asked he will know how to answer it.

Personalization: Paul can capture insights about the students’ personality and suggest better ways for them to learn. He can also adapt the learning options to the time the student has available: if a student says he or she only has 15 minutes to study on a given day, Paul will select the best learning option for that day.

The 7 onlearning principles: essentials for a successful onlearning project

1. Life-long Learning and Learning in Micro-moments

The idea of continuously learning throughout life makes more sense when you think of learning in micro-moments. “You can’t learn for the rest of your life at the pace of an MBA,” argues José Cláudio Securato, Saint Paul Business School’s CEO and founding partner. People now want immediate answers for problems that previously would have taken hours or even an entire day to solve. Just as important is the idea of learning in small periods along the day, like during the fifteen minutes you have while waiting for a meeting or while commuting. Learning should become a more fluid, organic process.

2. Personalization

Teaching in a standardized way was important in the past, when the biggest challenge in education was scalability. Today, we need to take a step further: personalization is now the key. Students have limited time and that time needs to be used in the most efficient way – a way that makes sense to each one of them. “If I only have fifteen minutes, I can’t afford to learn something that isn’t what I need, that won’t do anything for me. If it won’t solve my problem, I will switch to another app,” says Securato.

3. Micro-certifications

If students learn in micro-moments, why can’t they get a certification from micro-moments? Micro-certifications represent the idea that the students must be masters of their own learning, choosing their certifications according to their career needs and personal interests. Securato suggests the idea of a LEGO structure as a metaphor for learning. “That’s interesting because it allows the student to connect the pieces and set up the learning path that he or she wants,” says Securato. The big trend right now are the micro-certifications that offer the possibility of creating more advanced certifications when you add them up.

4. Learning Through Social Media

The idea of learning through a social network is not new: communities of practice have been doing it for decades. But the social media platforms that we have now can empower that concept in an unprecedented way. “We believe that the power of learning through social media is yet to be explored. We can put that together with the micro-moments and take learning into practice, because when you enter a community of practice, or a social network, you can exchange practical information with people who have more experience,” says Securato.

5. Learning Through Artificial Intelligence

Cognitive computing is based on two major groundings: the use of unstructured data and natural language processing. Why is that important? Students’ minds don’t work in a linear, structured way, so cognitive computing allows students to learn at their own pace and choose their own learning path (instead of following the order of chapters in a book, for example) Another point is that students can ask questions in their own natural language and the artificial intelligence will understand what they mean.

6. Blended Learning

Combining different methods of learning according to the needs and personality of the student is other important onlearning principle. On LIT, for example, students can acquire micro-certifications studying online and then earn an MBA from the institution by completing a few extra steps.

7. Return on Investment

Education can be very expensive, and the rate of return can be questionable. “Everyone knows that education offers return. What we want is a higher return on investment for students,” says Securato. “People are investing increasingly more in education, but generally speaking, their earnings are the same.” Democratizing access to education is an important factor for an onlearning project.


Twenty Years of Evolution That Have Transformed Education

José Cláudio Securato, CEO and FoundingPartner at Saint Paul Business School

Adriano Mussa, Academic & Artificial Intelligence Research Director and Partner at Saint Paul Business School





End of Comments