Twenty Years Later, Rolling Dots Part of Daily Lexicon

Topics: Technology
Featuring the work of:
1992 Hertz Fellow
Senior Director, AI Systems, Meta

After sending a text, we expect the rolling dots that indicate a recipient is preparing to respond. Yet the indicator was created long before texting became ubiquitous.

Twenty years ago, Hertz Fellow Sandeep Singhal was part of an IBM team that developed it while working on instant messaging for companies. 

“We worked long, crazy hours deep into the night,” said Singhal, currently a senior computer scientist with Facebook. “We would open up a conference call and literally just sit on the phone with each other, even though our offices were right next door.” The team needed the very collaboration tools they were in the process of developing. 

The IBM team had already developed messaging capability for video games. As team members grew more accustomed to using chat rooms and messaging platforms, they discovered new patterns, behaviors, and expectations. For example, they realized that once they sent a message, they had no idea whether it had been received. They’d sit alone, wondering if the recipient was taking a long time to respond to a message, or simply ignoring it.

“One night we began brainstorming about how to provide visual feedback that the message recipient had received the message and was getting ready to respond,” Singhal said. They decided to show the text response being typed in real time. 

“Our first attempt completely backfired,” he said. “Because putting a message in writing has a greater sense of permanency, people began to think about what they were going to write, type it somewhere, and make sure it was right before they sent it. They didn’t want to come across as unprofessional or unable to spell.”

The team realized that the sender didn’t want to know what the recipient was typing, just whether they were typing at all. The text indicator was born.

More sophisticated interactions

Now, 20 years later, Singhal sees an opportunity to move from flat screen collaboration into something more interactive, whether it uses 3-D, augmented reality, or virtual reality. 

The pandemic ushered in “an era where higher fidelity and more natural interaction while remote becomes more and more important,” Singhal said. “Collaboration patterns are becoming more complex, data-driven, and fluid.” 

Pulling together shared data or data sets in the cloud and fluidly moving from one piece of information to another is an increasing requirement for collaboration across multiple groups, said Singhal, who is an expert in storing, moving, and sharing massive amounts of data as the senior director of engineering for storage at Facebook. He previously led cloud storage at Google.

“I'm focused on figuring out how to store large data sets, keep that data safe, and make that data easy to manage,” he said. “I’m also focused on making data delivery fast for our employees and our customers.” 

Enabling group collaboration

Singhal’s positions at Facebook and Google have brought him full circle to the topic of his doctoral research at Stanford University, which focused on enabling large numbers of people to collaborate in a shared space. At the time, this had applications in gaming, military training systems, and scientific simulation. The landscape has matured, but his approach has remained the same.

“It’s all about taking a really hard problem and boiling it down as something you can roll out, deploy, and then learn from and iterate,” he said. 

Singhal continues to be as fascinated by computer science as he was when he took his first coding class at the age of 8. In particular, he’s most excited by developments in artificial intelligence (AI) and machine learning (ML), a subset of AI in which computer algorithms improve automatically through experience. 

“In the last five years, we’ve seen ML and AI go from being a curiosity, done in isolation by specialists, to becoming mainstream must-have technology in nearly every field,” Singhal said. As it becomes more and more mainstream, AI is forcing technological innovation in all areas of infrastructure as demand continues to increase.

The important thing is striking a balance between having a broad, long-term vision and figuring out how to get something done as quickly as possible and made available to users—like the dots that indicate someone’s typing, Singhal said. “You really want to get something built that people can actually use and experience.” 

Singhal is next contributing his expertise to help the Hertz Foundation. He has joined the network of Hertz Fellowship interviewers and will start meeting with first-round candidates in January 2021.