Executive Summary
- Cornell Tech researchers developed AI interfaces to enable individuals with speech disabilities to make timely jokes using AAC technology.
- The study introduced four AI-assisted interfaces – Full-auto, Keywords, Context Bubble Selection, and Wizard – offering varying degrees of control and speed in joke generation.
- Participants favored time-efficient interfaces after testing, even if the generated jokes were not entirely their own, highlighting the importance of timing in humor.
Event Overview
A team of researchers from Cornell Tech is exploring how AI can help people with speech disabilities make timely jokes during conversations using Augmentative and Alternative Communication (AAC) technology. The research addresses the challenge of slow communication pace with traditional AAC systems, which can hinder social interaction and humor. The study introduced four AI-assisted interfaces, each offering a different balance of speed and control in generating humorous comments.
Media Coverage Comparison
Source | Key Angle / Focus | Unique Details Mentioned | Tone |
---|---|---|---|
Cornell Chronicle | AI interfaces for humor in AAC users | Details on four AI interfaces: Full-auto, Keywords, Context Bubble Selection, Wizard. Mentions the Siegel PiTech Ph.D. Impact Fellowship and Amazon Cloud Computing Grant. | Positive and informative, highlighting the benefits and challenges of the technology. |
Key Details & Data Points
- What: Development of AI-powered interfaces to assist individuals with speech disabilities in making timely jokes during conversations.
- Who: Researchers from Cornell Tech and the Cornell Ann S. Bowers College of Computing and Information Science, including Ph.D. student Tobias Weinberg and assistant professor Thijs Roumen, along with assistant professor Stephanie Valencia of the University of Maryland. Participants with speech disabilities also played a key role.
- When: The research was presented at the Association for Computing Machinery’s Conference on Human Factors in Computing Systems (CHI ’25), April 26-May 1 in Yokohama, Japan.
- Where: Cornell Tech, New York City; University of Maryland; Yokohama, Japan (CHI conference).
Key Statistics:
- Key statistic 1: Best Paper Honorable Mention award and Jury Best Demo award at CHI ’25.
- Key statistic 2: Four AI-assisted interfaces were tested: Full-auto, Keywords, Context Bubble Selection, and Wizard.
Analysis & Context
The research highlights the importance of humor and self-expression for individuals with speech disabilities and how AI can bridge the communication gap. The four AI interfaces represent a spectrum of approaches, balancing agency and efficiency. The study's findings emphasize the significance of timing in humor, which is often overlooked in traditional AAC systems. This project has the potential to improve the quality of life for individuals with speech impairments by enhancing their social interactions and emotional connections.
Notable Quotes
Humor is a crucial form of social interaction and is usually taken as a trivial thing. But for someone with a speech impairment, it is not trivial to match the timing of the conversation.
So much work is focused on the direct ‘needs’ of people with disabilities, but empowering self-expression, humor and emotional connections is often overlooked. I hope we paved a path toward more research in this direction.
The time window where the humorous comment can land is so short. Even if the joke is not fully mine, it would still serve its purpose – which is to make other people laugh.
Conclusion
The Cornell Tech research demonstrates the potential of AI-powered interfaces to empower individuals with speech disabilities to participate more fully in social interactions through humor. The study's findings emphasize the importance of timing and the trade-off between agency and efficiency in designing these interfaces. Future research should focus on addressing challenges such as scrolling difficulties and improving the emotional nuance of text-to-speech voices to provide a more personalized and expressive communication experience.
Disclaimer: This article was generated by an AI system that synthesizes information from multiple news sources. While efforts are made to ensure accuracy and objectivity, reporting nuances, potential biases, or errors from original sources may be reflected. The information presented here is for informational purposes and should be verified with primary sources, especially for critical decisions.