Find A Way Initiative Member Solutions and Ideas. 

The reason this initiative is named "Find A Way" rather than 'Do This Thing' is due to the novelty and complexity of the issue at hand. Currently, there is no single solution for addressing it effectively. So, we must collectively endeavor, in good faith, to discover viable methods.

As time progresses, it is anticipated that certain strategies will emerge as more suitable for addressing the challenges presented by new technologies like AI. For now, our focus is on exploring new approaches and assessing their effectiveness in preventing discrimination against individuals with dyslexia. We will evaluate the evolution of these approaches transparently and collaboratively, using common and devloping metrics.


Current List of Issues:


1. Misuse of AI for Diagnosis:

   AI technology is being utilized by non-medical professionals to diagnose, expose, or discriminate against individuals with dyslexia, dysgraphia, dyspraxia, and various other learning disabilities. This discrimination manifests in several forms, including:

   

   1.1 Typing Analysis:

   Current AI technologies can analyze typed messages for indications that the writer has learning disabilities, such as dyslexia, dyscalculia, and dysgraphia, among others. This analysis can occur even if spelling errors have been corrected previously in word-processing software, as hints of some types of disability can be expressed simply in the way the text and the ideas are presented. Such technology leaves individuals vulnerable to discrimination in numerous forms of communication, including job applications, health insurance applications, work email correspondences, text messaging and social media posts.


   1.2 Handwriting Analysis:

   AI can also analyze handwritten documents to assess the likelihood of learning disabilities. This exposes individuals to discrimination when submitting written forms for job applications, insurance applications, and other official submissions. Even the way an address is written on a letter could potentially reveal a learning disability.


   1.3 Video Analysis:

   AI technology is used to analyze videos for micro-expressions and visual cues that may indicate learning disabilities. This can lead to discrimination against individuals in various scenarios, including job applications.


   1.4 Audio Analysis:

   Speech pattern analysis via AI can reveal indicators of dyslexia or other learning disabilities, posing a risk of discrimination during recorded interviews or public speeches.


2. Discrimination for Using AI Tools:

   Individuals with learning disabilities are often encouraged to use technology to support their needs. However, using AI for assistance, such as language models for writing aid, can lead to discrimination. Emails and social media messages may be flagged for AI use, and job applications might be screened out for suspected AI assistance.



Currently Attempted Solutions:


As of February 5th, 2024, several initiatives are underway to address these concerns. For instance, a company called "I Believe In You" has developed a GPT model based on OpenAI technology. This model is designed to understand and address the writing challenges faced by individuals with dyslexia. Importantly, it refuses to analyze texts for signs of dyslexia or to diagnose based on writing, demonstrating that it is possible to create controls around this technology without diminishing its supportive capabilities.