Me, Myself, and A.I.
Although I didn’t grow up with modern technology, I have always been fascinated by it. Growing up in Iraq, I remember feeling excited whenever our neighborhood’s 2-hour daily window of electricity came around. That meant I could plug in the Yamaha computer and start writing lines of code in Basic. The thrill of writing some words and seeing a machine turn them into a basic shape was unparalleled.
After I moved to the US as a refugee and accessed high-speed uncensored internet, I started learning about everything from cooking skills to coding skills. I quickly realized two things: everything is achievable with technology, and access to technology should not be limited. That’s why I decided to pursue a career in technology, to make sure everyone has equal access to it.
A lot of my work over the past 15 years has dealt with Machine Learning and Artificial Intelligence. I’m not an AI hipster by any means, but I’ve published about 10 peer-reviewed papers on the subject and have been working in the field closely enough to know that right now in 2023, things have changed.
Adaptive UIs for Older Adults
During my first Masters degree in Human-Centered Computing, I helped develop a browser extension that monitored the user’s cursor behavior and analyzed it in real-time using an A.I. model trained on data collected from people with Parkinson’s or tremors. If the model classified the user’s cursor movement as consistent with that of a person with mobility difficulties, it signaled that the user was having trouble interacting with the interface. I designed the interface to either alert the user of these difficulties or automatically adapt itself by changing the size of elements on the screen. This made it more efficient and easier to use.
You can see the results of this work here, here, and here.
Dytective: A video game to detect dyslexia
While I was a visiting scientist at Carnegie Mellon University, I co-invented Dytective with three other brilliant scientists (Luz Rello, Jeffrey Bigham, and Miguel Ballesteros). We used a dataset collected by Dr. Luz Rello of the types of spelling mistakes that people with dyslexia often made, trained an A.I. algorithm on it, and reverse-engineered it to build a web-based game targeted for children. The game had different levels where kids performed tasks like finding a single “p” in a grid of “q’s”. Based on how fast they did it and how many errors they made, the A.I. model would classify with ~90% accuracy how likely it was that the player had the risk of having dyslexia. This is one of my favorite projects ever because of its potential for impact — dyslexia is the leading cause for kids to drop out of school, and ~10% of the world’s population is dyslexic.
Dytective is still live to this day under the ChangeDyslexia.org non-profit and you can see my work on it here, here, here, here, and here.
Democratizing Interaction Design with AI
Between 2016 and 2020, I was able to complete my Ph.D. program in only four years, thanks to AI. This program would have normally taken about 6–7 years to complete. I did my Ph.D. thesis on Distributed Interaction Design, where I built a platform that enables technology creators to run co-design sessions in a distributed fashion. This means people who cannot have their voices, abilities, and preferences represented in how future technologies are made because they live somewhere far away from the creators or they have a disability that prevents them from doing so can now take part in how creators shape the way they are building new technologies. The big problem I faced when enabling researchers to do large scale co-design studies was how to analyze hundreds and thousands of designs from hundreds of participants efficiently. This is when I created Crowdsensus, a tool that uses crowdsourcing and unsupervised machine learning to analyze the results of these large-scale design studies.
You can see the results of this work here, here, here, and here.
What Now?
Now I’m working on an A.I. productivity tool designed for couples, called 3rd Brain. 3rd Brain provides couples with a shared space for tasks, events, and thoughts, helping them share the mental load of their daily lives and increase their collective productivity. It is also an AI tool that efficiently captures and organizes their thoughts and shares them with their partner, reducing the overhead of maintaining a productivity system for two.
Head over to 3rd-brain.com and to learn more. If you wanna chat about AI, design or anything else, feel free to reach out to me anywhere online @theaxali.
