Skip to main content

University of Michigan School of Information

Menu

Media Center

UMSI welcomes media inquiries

5 information experts look at where the trends are taking us

A screen with chat quotes.

Tuesday, 02/21/2023

The news is full of information breakthroughs and breakdowns: artificial intelligence, misinformation and “fake news,” and decision-making increasingly guided by algorithms. Five experts at the University of Michigan School of Information share trends they’re seeing in the information world and what they could mean for the rest of us.  

Kevyn Collins-Thompson, UMSI associate professor of information, director of the Master of Applied Data Science degree program

Kevyn Collins-Thompson’s interests include developing algorithms to connect people with information that helps them learn and discover. Products of Collins-Thompson’s work include intelligent tutoring systems that help improve children’s literacy skills, and search and recommender systems that connect students with the right resources at the right level of difficulty for their learning goals.

Collins-Thompson says people should get ready for lots more generative AI models like ChatGPT

“I think generative AI will have a huge impact on many different sectors of the economy. For example, the gaming industry is going to be revolutionized because before, you had to have a team of artists painstakingly create these graphical worlds. Now, they can be generated on the fly. 

As for learning, ChatGPT is based on large language models and it can produce very realistic, semi-intelligent conversation. It’s great at summarizing or translating information, and I know one researcher who believes that ChatGPT will be good at helping his PhD students generate first drafts of their papers because many of them struggle with technical writing. There are also downsides to this approach, but in general generative AI models do have the potential to save time in the early stages of a project.

In terms of concerns, I think people are correct to be skeptical about generative AI. It will sometimes generate misinformation. It can make mistakes. And since it’s trained on vast amounts of human language it definitely incorporates all the ugliness of human biases. There are risks associated with it, but if we pair generative AI with AI systems that can verify, fact check and identify sources, we could mitigate many of those concerns.” 

Nicole Ellison, UMSI Karl E Weick Collegiate Professor of Information

Nicole Ellison’s research studies how users self-identity and self present on social media, as well as how algorithms influence our identity and relationships with others. 

Ellison is looking forward to “fascinating conversations” about the relationship between humans and robots. 

“It’s going to be more and more interesting, and challenging, to ascertain the veracity of anything that’s not synchronous face to face communication. I think we’ll see some potentially fascinating conversations about how this shapes how we understand other people and how we understand our own identities. For instance, if you train a chatbot with fifty years of me speaking, ran it and had it answer some interview questions, is that me? How do we understand who the producer of that content is? And if the chatbot is able to hold a conversation with someone, how do we as humans understand that relationship? How do we think about who or what we’re communicating with?  

Steve Oney, UMSI associate professor of information

Steve Oney’s research is focused on accessibility and making programming easier for everyone. One of  Oney’s goals is to make it so that people who are not programmers can engage in the information world, make apps and collaborate with one another on cool ideas. 

Oney is excited about ChatGPT’s ability to make programming available to anyone. 

“I think the biggest news, beyond generating natural language answers, is that it also does a surprisingly good job of generating code. I don't think any software engineering jobs are at risk in the near-term future, as the code that it produces isn't too advanced and can also be buggy. Still, I think it's fairly clear that in the long term, this will change the way that we write code.

I think one of the most interesting immediate implications is that it lets nearly anyone write code. Even if a person doesn't understand how to read or write code, they can describe what they want, run it without reading the code, look at the output and describe the changes they want to make. With some improvements, future code generation tools like this could make some form of programming available to nearly anybody."

Kristin Fontichiaro, UMSI clinical professor of information

Kristin Fontichiaro works with libraries to deepen the learning experience of students, teachers and the general public. Currently, Fontichiaro works actively with public library directors to develop strategic communications amid book bans and attacks on library systems within schools and communities. 

Fontichiaro is concerned for the future of democracy. 

“Public libraries are in a really tough position because they’re places of public accommodations. The period that we’re in is scary and I’m seeing a lot of resonances to Nazi Germany in the 1930s. I’m hoping we can come out of this, that we’ll see more pushback against book bans. Libraries are important social hubs for a lot of people in the community. But we’re seeing a lot of budgets flattened, lowered investment, lowered support and lowered trust in many states, including Michigan. 

When we talk about trends, some of the trends in libraries right now are not positive. In schools and public libraries, we’re seeing the most changes. We have highly orchestrated campaigns that are demonizing schools and libraries, and it’s never been like this. It’s a strange time, and I think there’s this trend of ‘I want everything in the world exactly how I want it.’ And when I was growing up, there was something beautiful about civic institutions like schools and libraries being places where different people and ideas came together. 

Right now, we as a society are not interested in compromise. We’re not interested in saying, ‘I don’t like that material, but I do like this other piece, and as long as there’s something for everyone, it’s fine.’  Some of that, I think, comes from a shrinking middle class, a sense that a lot of families have lost their footing. And then we have a power class that tries to rile people up and say, ‘The real problem isn't that your minimum wage hasn’t been raised in 10 years. The real problem is there's a library book you don't like.’

One thing we have to do is help library professionals continue to protect the institution’s goal to meet everyone’s information needs, while bringing people together for meaningful conversation and social engagement.”

Yan Chen, UMSI Daniel Kahneman Collegiate Professor of Information

Yan Chen is a behavior economist who studies the intersection of labor and technology. She has authored research papers looking at the gig economy, which makes up more than a third of the U.S labor force. 

Chen is looking forward to seeing companies like Uber, Lyft and TaskRabbit retain workers by increasing team identity and incentives.

“People are assigned tasks via algorithms, and there’s typically a lack of interaction between workers. They’re without any organization or friends or community. Because of this, 60% of Uber drivers leave within their first six months. Recruitment and training is costly, so providing incentives is important. We did an experiment with a large ride-sharing platform to increase engagement and bonds among drivers through incentive programs and by gamifying the work. And we found that after the team contest was over, workers continued to be more productive and engaged. 

I think in terms of trends we’re going to be seeing a lot more shipping of algorithmic production and using the algorithms to run team contests. We’re going to see a lot more companies implementing these contests, through algorithms, into their organizational structure. 

Because anecdotally, in the interview samples, the drivers say ‘I didn’t know anyone in the city. I was driving by myself and the best part of the teams and these games is meeting people. This gamification is already happening, and I think we’ll be seeing more of it and more of these information interventions.”

- Noor Hindi, UMSI public relations representative