There is no magic bullet for anything, tradeoffs are everywhere. There is no perfect solution or algorithm. You will experience a tradeoff in some form. For example, for our website we are constantly trading off on how much to modify a sentence versus how much fluency to maintain. The main goal should be to find balance, not to try and solve everything.
As part of my series on “Bleeding edge” technological breakthroughs that seem copied from science fiction I had the pleasure of interviewing Rohan Gupta. Rohan is the COO of QuillBot, a startup developing cutting-edge paraphrasing technology. He completed both his masters in computer science and his undergraduate degree in finance from the University of Illinois at Urbana-Champaign.
Thank you so much for doing this with us! Can you tell us a story about what brought you to this specific career path?
I’ve always been entrepreneurial. I remember selling silk pillow cases my mom brought from India when I was 10. I’ve also been fascinated by technology, and in my early high school years I used to program utilities and sell them online. When I went to college, I became obsessed with artificial intelligence, studying AI both in school and in my free time. When my close friend made a breakthrough in a paraphrasing model, enabling it to become state-of-the-art, I knew that I wanted to play a part. After that, we founded QuillBot together.
Can you share the most interesting story that happened to you since you began your career?
The most interesting thing has been talking to our users and seeing all the creative ways they have utilized QuillBot. One time, we talked to a high school teacher who was really struggling to write recommendation letters for students. She complained that all of them sounded the exact same, even though she knew her students were relying on her. She mentioned to me that QuillBot was able to help her rewrite recommendation letters tailoring each one to their respective student with a unique voice.
Can you tell us about the “Bleeding edge” technological breakthroughs that you are working on? How do you think that will help people?
QuillBot is a state-of-the-art paraphrasing tool. It uses machine learning to rephrase entire sentences while maintaining their original meaning. We released this technology in the form of a full sentence thesaurus. Users simply go to the site, input a sentence or paragraph, and watch QuillBot reword and restructure it. Users can then click on any word or phrase to have a drop-down thesaurus appear where they can make additional edits. We then use this data in order to train the AI and improve QuillBot. In the end, we want to achieve human-level proficiency in paraphrasing. QuillBot gives users creative suggestions on how to word and structure sentences. It can also fix grammar and improve fluency.
We are also working on linguistic style transfer. Basically, we want to be able to take any sentence and rewrite it in a particular style, tone, or reading level. For example, take a sentence and rewrite it as if it was written in old english or slang. We even want to be able to take a sentence and rewrite it so it’s simpler or more sophisticated. A potential use case of this technology would be adaptive learning in education. We could morph a story as the student is reading it to match their reading comprehension level — all while maintaining the meaning and flow of the story.
How do you think this might change the world?
We believe this technology will further blur lines between artificial intelligence and humans. By allowing the AI to be able to express its intentions in a wide range of ways, we can enable the AI to sound more human. In addition, we believe the QuillBot model will allow AIs to adopt many linguistic styles, enabling one AI to personalize the same communication to a diverse audience. Thus, it can enable virtual assistants and other AIs to become far more relatable. Our technology can also be used to increase the proficiency of many AI systems like translation, which can also have lasting effects on the world by helping to bridge linguistic gaps across different groups of people.
Keeping “Black Mirror” in mind can you see any potential drawbacks about this technology that people should think more deeply about?
There are a lot of frightening implications of this technology. Primarily, it makes one question who owns an idea. If an AI can rewrite a particular idea in a large amount of different ways, it would make it quite easy to steal other people’s work. This issue manifests itself in academia as plagiarism, but it can also be seen in blogging, writing, and reporting. Obviously, we do not condone the use of the technology for plagiarism, but technology tends to be disruptive and a double-edged sword. I feel like the wider dangers of this technology are how it can be used to generate fake news and astroturfing. By enabling AIs to communicate an idea in many ways, one can artificially create a sense of authority. If someone wanted to promote a fake news article, they could have the AI rewrite it 20 times, and then publish it across many platforms. People who then Google the topic will see a lot of supporting articles which may give them the impression that the story has more support than it actually does. For example, after the 2016 election, you can visibly see the impact that fake comments had on the opinions of others. An AI that could rewrite comments could further strengthen these misinformation campaigns.
Was there a “tipping point” that led you to this breakthrough? Can you tell us that story?
Originally, my cofounder was working on an AI project that would generate multiple choice questions for quizzes automatically after reading the relevant source material. In that process, he was looking for the AI to paraphrase the same question in multiple different ways to keep each quiz unique. He made a breakthrough that led the model to become state-of-the-art in paraphrasing. I joined the team, we made a quick pivot, and QuillBot was born.
What do you need to lead this technology to widespread adoption?
The biggest thing that we need to do is to increase the proficiency of the AI. The technology is good on an average case scenario, but we need to reduce its error rate while still increasing the difference between the modified and original sentences. Currently, for most cases, a human needs to approve Quillbot’s output. We feel that when the underlying technology is capable enough to generate sentences that do not require human review, Quillbot will disrupt many industries and start to become widely adopted.
What have you been doing to publicize this idea? Have you been using any innovative marketing strategies?
One of the main and innovative ways we promote QuillBot is by making unique content on our blog and sharing it. For example, we made this great data visualization of the growth of reddit, and it went viral. That visualization drove a lot of traffic to our website. We also post about QuillBot on various subreddits and try to spread the word that way. It’s a great platform to reach a diverse group of people while still maintaining relevancy. Finally, a lot of people find out about QuillBot through word-of-mouth referrals as well as through Google searches.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?
My father was a big inspiration to me. I always wondered why we would believe in my harebrained schemes and let me experiment freely. Most parents tell their children to follow conventional career paths. However, my dad is not a conventional parent. He always told me that I needed to learn how to fail. He told me that I should get an F on a test instead of getting constant As so I could learn how to deal with failure. See, he is an entrepreneur also, and he has suffered many failures while starting and scaling his own business. From him, I learned that it is not enough to take the first step, but rather that I need to keep on taking steps even after I fall down. Because of my dad, I have learned that I need to fail my way to success.
How have you used your success to bring goodness to the world?
We are not quite at the point where I would consider us to be a big success; however, we are always trying to share our experiences with those at an earlier stage than us. Anytime we have the chance, we share our advice with other entrepreneurs. Hopefully that will help some other people succeed. I believe that advice should always be paid forward. We have received valuable advice along the way, and we hope that we can be that helping hand to others. When entrepreneurs are inspired, and given the proper guidance, progress around the world is made. If someone with a great idea cannot materialize it, the world is the one that really loses out.
What are your “5 Things I Wish Someone Told Me Before I Started” and why. (Please share a story or example for each.)
You are a person of great influence. If you could inspire a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger. 🙂
My whole team heavily believes in the democratization of education. My cofounder is a college dropout and one of the most intelligent people I have ever met. His lack of a degree could impact his ability to get a job despite his immense qualifications. In fact, we are all self-taught programmers who consider the internet our best teacher. I believe removing the stigma of an online degree as well as reigning in what I refer to as “academic inflation” could have a major impact. What we are seeing is this misguided belief that a college degree is the bare minimum required for any job, and higher level degrees or degrees from prestigious organizations are the only ways to truly cement your qualifications — an “inflation” of educational standards. Rather, we believe that online degrees could enable people to develop highly technical skills on their own time and at their own pace, all while not having to pay exorbitant college tuitions. By opening up more opportunities to a larger number of people, the democratization of education could dramatically increase the number of productive members of society, improving the lives of everyone.
Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?
“There is no man wise enough to learn from the experiences of others” -Voltaire.
To me, this quote means that you have to learn things for yourself, and that the best way to do that is by constantly experimenting and putting yourself out there. The impact of a true life lesson is far greater than a few words of wisdom. It was this sentiment that pushed me to become an entrepreneur, I live life like a constant experiment knowing failures are stepping stones that provide me valuable life lessons and help me grow.
How can our readers follow you on social media?
Thank you so much for joining us. This was very inspirational.