Study after study shows that walking is very good for those who are able, and . A 2023 study found that even 4,000 steps a day . (The .) For each 1,000 extra daily steps, risk decreased by 15%.

have been used since 2007 to quantify how quickly people can typically walk to amenities like grocery stores and schools in an area. ; for instance, Seattle’s 74 means it’s “very walkable.” It may seem self-evident that in cities and towns with better scores people tend to walk more. But it’s surprisingly difficult to tease out the cause and effect: Do walkable cities prompt people to take more steps, or do people who want to walk tend to live in more pedestrian-friendly cities?

New research led by the 91̽�� provides clear evidence that highly walkable areas lead to significantly more walking. Using data from the step-tracking app, authors compared the steps per day of 5,424 people who moved one or more times among 1,609 cities in the United States. Across all relocations, when the Walk Score rose or fell more than 48 points, average steps increased or decreased by about 1,100 per day. But when people moved between similarly walkable cities, their steps stayed about the same. These findings held across people of different ages, genders and body mass indexes.

For instance, the study tracked 178 people who moved to New York City (Walk Score 89) from different cities with an average score of 48. This group’s average daily steps rose by 1,400 upon moving to New York, from 5,600 to 7,000. Moves from New York to less walkable cities showed the inverse: People averaged 1,400 fewer steps.

The authors Aug. 13 in Nature.

“Some of suggested that our physical, built environment makes a big difference in how much we move, but we couldn’t produce particularly strong evidence showing that was the case,” said lead author , a 91̽��associate professor in the Paul G. Allen School of Computer Science & Engineering. “The large data set we worked with for this new study gave us a unique opportunity to produce this strong, compelling evidence that our built environments do indeed causally impact how much we walk.”

Working with an anonymized data set from 2.1 million people who used the Argus app between 2013 and 2016, the team pulled a subset who had moved and stayed in their new location for at least three months. They normalized for demographics and changes in seasons. They also filtered out days with fewer than 500 steps or more than 50,000, as well as days around moves.

The greatest change in walking the study observed was in the moderate intensity range (100 to 130 steps per minute). Moves that increased Walk Scores more than 49 points were associated with twice as many subjects recording at least , the recommended minimum.

Althoff stressed that while the study provides the strongest evidence to date, no data set is truly representative of the whole U.S. population. For instance, the subjects in this study had all downloaded a step-counting app, which can affect results.

“Our study shows that how much you walk is not just a question of motivation,” Althoff said. “There are many things that affect daily steps, and the built environment is clearly one of them. There’s tremendous value to shared public infrastructure that can really make healthy behaviors like walking available to almost everybody, and it’s worth investing in that infrastructure.”

Other co-authors on this paper include of NVIDIA Research and , , and of Stanford University.

This research was funded in part by the National Institutes of Health, the National Science Foundation and the Gates Foundation.

For more information, contact Althoff at althoff@cs.washington.edu.

Empathy is critical to having supportive conversations about mental health. But this skill can be tricky to learn, especially in the moment when a person is sharing something hard.

A team led by researchers at the 91̽�� studied how artificial intelligence could help people on the platform , where people give each other mental health support. The researchers developed an AI system that suggested changes to participants’ responses to make them more empathetic. The system helped people communicate empathy more effectively than traditional training did. In fact, the best responses resulted from a collaboration between AI and people.

The researchers Jan. 23 in Nature Machine Intelligence.

91̽��News reached out to senior author , 91̽��assistant professor in the Paul G. Allen School of Computer Science & Engineering, for details about the study and the concept of AI and empathy.

Why did you choose the TalkLife platform to study?

Tim Althoff: Prior research suggests that peer-support platforms could have a significant positive impact on mental health care because they help address the massive challenge of access. Because of insurance issues, stigma or isolation, many people find that free online peer-support platforms are all they have access to. TalkLife is the biggest peer-support platform globally and it has a large number of motivated peer supporters.

Also, TalkLife leadership recognized the importance and potential impact of our research on how computing can empower peer support. They kindly supported our research through collaboration, feedback, participant recruiting and data-sharing.

What inspired you to help people communicate with more empathy? ��

TA: It is well established that empathy is critical for helping people feel supported and for forming trusted relationships. But empathy is also complex and nuanced. It can be challenging for people to find the right words in the moment.

While counselors and therapists are trained in this skill, our prior research established that . We also found that peer supporters do not learn how to express empathy more effectively over time, which suggests that they could benefit from empathy training and feedback.

On the surface it seems counterintuitive to have AI help with something like empathy. Can you talk about why this is a good problem for AI to solve?

TA: What the AI feedback can do is be very specific and be “contextual” and give suggestions about concretely responding to a message that’s right in front of someone. It can give someone ideas in a “personalized” way rather than through generic training examples or with rules that may not apply to every single situation a person will face. It also only pops up if someone needs it — if their response is great, the system can give a light touch of positive feedback.

People might wonder “why use AI” for this aspect of human connection. In fact, we designed the system from the ground up not to take away from this meaningful person-person interaction. For example, we only show feedback when needed and we train the model to make the smallest possible changes to a response to communicate empathy more effectively.

How do you train an AI to “know” empathy?

TA: We worked with two clinical psychologists, at Stanford University and in the 91̽��School of Medicine, to understand the research behind empathy and adapt existing empathy scales to the asynchronous, text-based setting of online support on TalkLife. Then we had people annotate 10,000 TalkLife responses for various aspects of empathy to develop AI models that can measure the level of expressed empathy in text.

To teach the AI to give actionable feedback and concrete suggestions, we developed a reinforcement learning-based system. These systems need a lot of data to be trained, and while empathy isn’t expressed as often as we would like on platforms such as TalkLife, we still found thousands of good examples. Our system learns from these to generate helpful empathy feedback.

In your evaluation of this system, did you see people becoming reliant on AI for empathy or did people learn how to be more empathetic over time?

TA: Our randomized trial demonstrated that peer supporters with access to feedback expressed between 20% and 40% more empathy than supporters in the control group that did not have access to such feedback.

Among our participants, 69% of peer supporters reported that they feel more confident at writing supportive responses after this study, indicating increased self-efficacy.

We further studied how participants made use of the feedback and found that peer supporters did not become overly reliant on the AI. For example, they would use the feedback indirectly as a broader inspiration rather than “blindly” following the recommendations. They also flagged feedback in the few cases when it was not helpful or even inappropriate. I was excited that the collaboration between human peer supporters and AI systems led to better outcomes than either one on their own.

I also want to highlight the significant efforts we took to consider and address ethical and safety risks. Those include having the AI work with the peer supporter instead of the person currently in crisis, conducting the study in a TalkLife-like environment that is intentionally not integrated into the TalkLife platform, giving all participants access to a crisis hotline and allowing peer supporters to flag feedback for review.

What do these results mean in terms of the future of human-AI collaboration?

TA: One area of human-AI collaboration that I am particularly excited about is AI-supported communication. There are so many challenging communication tasks with critical outcomes — from helping someone feel better to challenging misinformation on social media — where we seem to expect people to do well, without any form of training or support. In most cases, all we are given is an empty chat box.

We can do better, and I believe that natural language processing technology can play a big role in helping people achieve their conversational goals. Specifically, our study shows that human-AI collaboration can be effective even for complex and open-ended tasks such as having empathetic conversations.

Additional co-authors on this paper are and , both 91̽��doctoral students in the Allen School; , an affiliate professor in the psychiatry and behavioral sciences department in the 91̽��School of Medicine and the CEO of Lyssn.io, Inc.; and at Stanford University. This research was funded by the National Science Foundation, the National Institutes of Health, the Bill & Melinda Gates Foundation, the Office of Naval Research, a Microsoft AI for Accessibility grant, a Garvey Institute Innovation grant, the National Center for Advancing Translational Science, Clinical and Translational Science awards and the Stanford Human-Centered AI Institute.

For more information, contact Althoff at althoff@cs.washington.edu.

Grant numbers: NSF grant IIS-1901386, NSF grant CNS-2025022, NIH grant R01MH125179, INV-004841, N00014-21-1-2154, KL2TR001083, UL1TR001085 and K02 AA023814