Exploring outlooks towards generative AI-based assistive technologies for people with Autism
Project Overview
The document explores the role of generative AI in education, emphasizing its potential applications and implications, particularly for individuals with autism. It highlights the use of deepfake technology, specifically Nvidia's eye contact functionality, as a tool to assist autistic individuals in improving communication during video conferencing. While some members of the autistic community view this technology as beneficial for facilitating interaction, others express concerns that it may reinforce societal pressures related to eye contact, potentially leading to harmful masking behaviors. The mixed sentiments underscore the complexity of integrating generative AI in educational contexts. The findings indicate that while generative AI can enhance human-computer interaction and communication, there is a critical need for further research to fully understand its implications and ensure that such technologies are designed and implemented thoughtfully. The document ultimately calls for a balanced approach to deploying generative AI in educational settings, taking into account the diverse perspectives and needs of users to maximize positive outcomes while minimizing potential drawbacks.
Key Applications
Nvidia's eye contact functionality in video conferencing
Context: Assistive technology for individuals with autism during online meetings.
Implementation: The technology modifies live video to simulate eye contact, allowing individuals to appear as if they are making direct eye contact with viewers.
Outcomes: Some autistic individuals found it could help them communicate more effectively in online settings, while others expressed concern about its implications on social norms.
Challenges: Mixed reactions regarding the effectiveness and appropriateness of the technology, with many feeling it could be a harmful masking strategy.
Implementation Barriers
Social Norms Barrier
Generative AI technologies may reinforce neurotypical communication norms, such as the expectation of making eye contact.
Proposed Solutions: Encouraging awareness and acceptance of diverse communication styles and challenging the necessity of eye contact in social interactions.
Usability Barrier
Concerns about the technology's usability in various workstation setups and its compatibility with users' natural behaviors.
Proposed Solutions: Conducting human factors research to understand how autistic individuals prefer to interact with such technologies.
Perception Barrier
The unsettling nature of the technology, which some users find creepy or unnatural.
Proposed Solutions: Investigating the uncanny valley effect and improving user experiences through design and feedback.
Project Team
Deepak Giri
Researcher
Erin Brady
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Deepak Giri, Erin Brady
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai