Key Takeaways From “AI for Good: A Conversation With Josh Tyrangiel”
Watch a recording of the event here.
On June 9, 2026, the Aspen Economic Strategy Group hosted author Josh Tyrangiel to discuss his new book, AI for Good: How Real People are Using Artificial Intelligence to Fix Things That Matter.
The book explores real-world examples in areas such as education, healthcare, and public services where people are solving challenges in their field through the use of AI. Tyrangiel argues that, while this technology carries risks, society cannot afford to reject AI out of fear. Rather, we must work to shape its applications in ways that benefit society.
Tyrangiel suggests that AI integration is most successful when certain conditions are met. First, AI systems must augment, rather than replace, human judgment and activity. Second, AI systems must be designed to allow for oversight of output — a condition he refers to as “keeping humans in the loop.”
Finally, he points out that professionals operating on the front lines of their fields, not technical AI experts, are best positioned to identify existing inefficiencies and create solutions. The book’s case studies feature teachers, doctors, and government employees, many of whom do not have a formal tech background, who are successfully solving challenges within their fields with the assistance of AI.
Tyrangiel opened the program by identifying what the book does not set out to accomplish — namely, presenting AI as a riskless endeavor or suggesting that the technology is a “silver bullet” that will eliminate long-standing issues across policy sectors. Rather, he hopes that readers will work to acknowledge and mitigate risk where it exists while remaining imaginative about how society can use AI to address challenges that exist today.
AESG Director Melissa S. Kearney questioned if there are any contexts where AI doesn’t belong, citing the lack of success in applying technology to improve outcomes in education. Tyrangiel acknowledged that the technology may have greater potential to directly help teachers rather than students, pointing to examples of instructors using AI-infused technology to adjust their lesson plans to better reflect how today’s students are interacting with the world.
Kearney suggested that the AI integration Tyrangiel described throughout much of the book is evolutionary — rather than revolutionary — in nature. In response, Tyrangiel argued that the perception that AI use has to be revolutionary is pushed by labs whose primary motivation is to market their products as the most competitive offering on the market.
Despite prevalent advertising that positions AI as transformative, Tyrangiel maintained that today’s technology still has substantial limitations. Namely, large language models are trained on an enormous amount of historical data but cannot keep pace with the speed at which humans and their culture change. He asserted, “You can move really quickly. You can get lots of processing power. But what AI really doesn’t know yet is what do we want and how does it change?”
Tyrangiel also suggested that AI integration fails when humans are removed from the loop. He pointed to a sepsis identification algorithm that was recently implemented in Cleveland Clinic hospitals as evidence — medical professionals gained the most insight from the system only when they were able to compare the system’s diagnosis with their own knowledge and experience treating the condition.
Kearney closed by asking how individuals can work to advance an “AI for Good” agenda. Tyrangiel encouraged attendees to mobilize and demand the AI systems they want to see — otherwise, the technology will be shaped by those who do not prioritize using AI for the betterment of society. He concluded, “If you are passive in the face of this wave of technology… you’re going to get the very worst of AI.”