AI is becoming increasingly ubiquitous with the rise of ChatGPT, Github Copilot, DALL-E and so on. However, it is a well-established fact that algorithms that model social outcomes are not fair. Knowing why this is so, and what you can do about it, can help us make better decisions and evaluate outcomes more holistically when these technologies are used on us.
WIIFY
You can make better decisions by knowing how AI may be flawed
Key points
AI can never be completely fair
As users, we should double-check and question findings in AI
As developers, we should communicate these limitations to users
Impact
Know: How and why AI is biased
Believe: AI is not fair, and knowing this can help us make better decisions
Do: Look out for how AI technologies are developed and for developers, communicate the limitations of AI
AI is becoming increasingly ubiquitous with the rise of ChatGPT, Github Copilot, DALL-E and so on. However, it is a well-established fact that algorithms that model social outcomes are not fair. Knowing why this is so, and what you can do about it, can help us make better decisions and evaluate outcomes more holistically when these technologies are used on us.
WIIFY
You can make better decisions by knowing how AI may be flawed
Key points
Impact
Slides
Slides download link