Junior Vedh Krishnan recently traveled to California to share his newly-developed app at the AI in Education Summit hosted by the Khan Lab School. 

“The fact that I got to visit Silicon Valley for the first time and finally do what I’ve always wanted to do was amazing: [I got to] see this really influential hub of technology,” said Krishnan. 

While at the event, Krishnan spoke to over 300+ attendees, Stanford graduate students, and the founder of Khan Academy himself: Sal Khan.

“There were definitely times where I felt overwhelmed. I thought, ‘Oh my god, the creator, Sal Khan himself, is actually here and he’s hearing me speak. What if I mess up?’” said Krishnan. “I tried to tell myself ‘let’s not overthink this; let me just be myself.’”

At the summit, Krishnan presented GigianoAI, an AI app he developed his sophomore year that utilizes lectures from history teacher Mr. Thomas Gigiano to provide personalized help and feedback to AP World and European History students. 

“People ask: ‘Vedh, are you trying to eliminate the teacher?’ No, I’m trying to strengthen the connection between the teacher and the student,” said Krishnan. 

The idea emerged after Krishnan noticed students struggling to find specific answers using lecture videos Mr. Gigiano created during COVID. 

“Whenever we’d be confused after a long lesson that Mr. Gigiano would give, [he] would tell us to watch his lectures…and they were really long,” said Krishnan. “People knew around 70% of the content; it was just gaps. There were certain specific questions they needed [answered].”

As a result, Krishnan, a second-year computer science student at the time, turned to his coding skills. 

“[My computer science class at LFHS] was the first time I was able to type some actual code… that was when AI was starting to emerge, and I was seeing the potential of AI relative to programming,” said Krishnan.

So, how hard is it to create a program like this? Computer science teacher Steve Aronson explains that the difficulty is having the parts of the app communicate and training the tool to function properly.

“The front end is what the user talks to. The back end is where you store all the data. He’s had to work on those to have the two talk to each other, and each part requires its own problems [to be solved],” said Mr. Aronson. “Then there’s also training the system…He had to upload [the videos], store the results, and then basically turn [the videos] into that teacher.”

Creating and training GigianoAI took a significant amount of dedication and effort on Krishnan’s part, who was assisted by Mr. Gigiano himself.

“Mr. Gigiano was my biggest advocate and was really interested in this. After class, I would talk with him and make him test out the AI as we went. Around a month or so, we got a fully working app,” said Krishnan.

According to Mr. Gigiano, Krishnan ensured that the app contained specific guidelines so it was used solely as a helpful resource instead of a cheat tool to write full essays. Mr. Gigiano’s lectures contain an abundance of information–from specific historical events to how to write AP history essays–so Krishnan wanted to guarantee that students had to engage with the AI to understand the material by asking numerous questions.

Let’s say, for example, a student wanted to learn about Sir Issac Newton, as seen in Krishnan’s GigianoAI demonstration video. The student would pick that subject from a dropdown menu of all the lectures, and the AI would respond with a brief overview of the lecture. The student can then ask more questions on specific topics they are confused about; this allows them to obtain detailed answers and expand their knowledge. 

“He’s the guy who did all the groundwork…He wanted to make it so that the student had to continuously prod the actual AI to get to the answer: it was not too quick, and the AI would not write a full LEQ,” said Mr. Gigiano.

These parameters were set to combat rising concerns that AI tools are utilized by students to cheat rather than guide their learning. Addressing these worries, Mr. Gigiano asserts that the option for students to cheat has always been there: AI is just a new way to do so and needs to be checked through trackers. 

“‘I’m going to use AI to write a paper.’ People are worried about that, right? They’re worried that we can’t get research papers anymore because they can [use] AI,” said Mr. Gigiano. “Well, they could have just copied out of a book before. It’s the same problem that we’ve always had.”

Instead of focusing on the potential negatives, Mr. Gigiano believes that apps like Krishnan’s GigianoAI offer a massive benefit to students. 

“AI has the capability of scanning large, vast amounts of articles and data and then providing you with the most accurate research out there in one fell swoop, and I think that that can be very, very useful. It’s a communication revolution,” said Mr. Gigiano.

Krishnan was grateful for the experience and for sharing his thoughts, especially in a time when concerns about AI are being expressed. 

“Despite all this AI [discussion], the human aspect of it is really important in the entire process [of coding], and that’s something that I feel is a misconception,” said Krishnan.

Coming back from the trip, Krishnan is focused on having the app be accessible to students. It is fully functioning and ready to use; he just needs to complete discussions with the school about how the app will be funded. In the meantime, Krishnan is currently creating a new program–Build a Teacher–that expands help to students across all courses.

“What we have with GigianoAI is really special,” said Krishnan. “So many students are really enthusiastic about this, and we’re really excited to move this forward.”