To what extent can an API-based large language model (LLM) designed for autistic college students improve career readiness to aid in the school-to-workplace transition through biometric devices?
Abstract
Approximately 85-90% of autistic adults face employment challenges due to difficulties with workplace norms, hiring processes, and self-advocacy. Traditional career readiness programs often fail to accommodate the unique cognitive and social needs of autistic individuals. Large language models (LLMs), such as OpenAI’s GPT-4, present an opportunity to bridge this gap by providing accessible, judgment-free learning environments. However, existing LLMs are limited by neurotypical biases, generalized responses, and lack of adaptability to autistic users' needs. This study investigates the extent to which an API-based LLM, tailored for autistic college students, can improve social cue interpretation, self-perception, and career readiness during the school-to-work transition. A modified chatbot, Auti, was developed using GPT-4 with prompt engineering and API integrations to enhance accessibility and personalization. The study employs a mixed-methods approach, recruiting 24 college students (both neurodivergent and neurotypical) to compare interactions with modified and unmodified LLMs. Pre- and post-surveys, biometric feedback, and qualitative interviews assess participants’ experiences. Initial pilot results suggest that autistic users benefit from structured, simplified, and neurodivergent-friendly responses, enhancing their career readiness and workplace communication skills. This research contributes to the growing field of AI-driven career support for neurodivergent individuals and highlights the need for bias-free, adaptive LLMs to ensure equitable access to employment opportunities.
To what extent can an API-based large language model (LLM) designed for autistic college students improve career readiness to aid in the school-to-workplace transition through biometric devices?
TBD
Approximately 85-90% of autistic adults face employment challenges due to difficulties with workplace norms, hiring processes, and self-advocacy. Traditional career readiness programs often fail to accommodate the unique cognitive and social needs of autistic individuals. Large language models (LLMs), such as OpenAI’s GPT-4, present an opportunity to bridge this gap by providing accessible, judgment-free learning environments. However, existing LLMs are limited by neurotypical biases, generalized responses, and lack of adaptability to autistic users' needs. This study investigates the extent to which an API-based LLM, tailored for autistic college students, can improve social cue interpretation, self-perception, and career readiness during the school-to-work transition. A modified chatbot, Auti, was developed using GPT-4 with prompt engineering and API integrations to enhance accessibility and personalization. The study employs a mixed-methods approach, recruiting 24 college students (both neurodivergent and neurotypical) to compare interactions with modified and unmodified LLMs. Pre- and post-surveys, biometric feedback, and qualitative interviews assess participants’ experiences. Initial pilot results suggest that autistic users benefit from structured, simplified, and neurodivergent-friendly responses, enhancing their career readiness and workplace communication skills. This research contributes to the growing field of AI-driven career support for neurodivergent individuals and highlights the need for bias-free, adaptive LLMs to ensure equitable access to employment opportunities.
