Apply the inclusiveness principle
Inclusiveness in AI means ensuring that AI systems are fair, accessible, and are applicable to everyone. Inclusivity happens when you consider the diverse range of experiences from all people, communities, and geographies, especially the areas of society that are historically underrepresented.
AI systems that uphold inclusiveness in education must work well for diverse learner groups and not disadvantage anyone. They must be accessible to everyone easily, including people with disabilities. They must allow all learners to benefit equally.
Examples of inclusive AI systems include:
- facial recognition that works across all skin tones, ages, and genders.
- Interfaces that support screen readers for the visually impaired.
- Language translation that supports small regional dialects.
Microsoft's responsible AI standard requires that everyone can access AI systems, regardless of any disability. Responsible AI solutions must enable full global inclusion by offering alternative modes of interaction such as voice control, captions, and screen readers.
Many AI systems contain historical societal biases, such as facial recognition that struggles when identifying women or people of color, automated recruiting tools that prefer majority demographics, and product features that implicitly encode existing biases. When implementing AI systems, you should acknowledge these issues and take steps to address them.
You can have learners of diverse backgrounds and abilities test the AI systems to determine whether they can be used as intended by the broadest possible audience. You can also use common accessibility standards to assess your AI system and ensure it is accessible to people of all abilities.
It is important to speak up against exclusionary practices, advocate for greater diversity in teams implementing AI systems, consider the impact on the less privileged, and hold yourself and your teams accountable to inclusive principles.
Use cases of AI improving accessibility
Accessibility solutions like speech transcription and captioning, content readers, translation services, voice assistants, and computer vision create more inclusive learning experiences. You can take advantage of inclusive tools to improve learning experiences as shown below:
Conversation, meetings, and live broadcasting: Speech-to-text transcription and captioning make it easier for learners who are hard of hearing, learners with ADHD, learners who are multi-tasking, and learners in noisy or low-bandwidth environments to participate in online learning sessions.
Content reading: Text-to-speech technology like Read Aloud reads webpages, documents, and emails aloud, making it easier for learners with visual impairments, dyslexia, ADHD, or autism to access and consume digital learning content. It also enables all users to multi-task and listen to content on the go.
Communication: Speech-to-speech, text-to-speech, and optical character recognition (OCR) technologies enable communication across different languages and formats. By pairing Microsoft Teams with Microsoft Translator, learners and trainers who speak different languages can communicate in person or virtually using their preferred languages. Multiple learners and trainers can join a conference call, and each can communicate in their own language.
Device access: Facial recognition allows learners to log in to their learning devices with their face, making it easier for those with mobility issues to access their devices and log in hands-free.
Device control: Speech-to-text technology can allow learners of all abilities to control the learning device and access information and services through voice commands.
Images and surroundings: Image captioning provides descriptive text for images to make them accessible to learners with visual impairments. Image descriptions also improve search engine optimization (SEO) and provide alternatives for learners listening to content or learners who are in low-bandwidth environments.
The next unit covers the principle of transparency. You will learn the value of AI is rooted not only in its accessibility, but also in its comprehensibility. Transparency serves as the bridge that connects AI creators with users, nurturing a relationship built on trust and understanding. It compels creators to clarify the complexities of AI and provide explanations of how their systems operate. With AI systems becoming increasingly integral to our daily lives, transparency is important as it ensures that everyone can interact with AI with confidence.