AutoRob #ScholarStrike Presentation - That Ain't Right: AI Mistakes and Black Lives

Summary

The presentation addresses the impact of AI errors on marginalized communities, using the case of Robert Williams's wrongful arrest due to facial recognition mistakes. It discusses the broader implications of AI in social justice, the need for diversity in AI research to prevent biased outcomes, and the importance of empathy in technology development. The speaker calls for systemic change, highlighting the underrepresentation of African Americans and Latinos in tech and advocating for equal opportunities as mandated by civil rights acts.

Takeaways

• 😀 The presentation discusses the impact of AI mistakes on Black lives and society, emphasizing the importance of addressing systemic issues in AI development.
• 🔍 The speaker highlights the case of Robert Williams, who was falsely arrested due to a facial recognition error, illustrating the real-world consequences of AI inaccuracies.
• 💻 The presentation calls for a more diverse and inclusive approach to AI research and development to prevent biases and improve outcomes for all communities.
• 🌐 It mentions the role of organizations like blackincomputing.org in advocating for change and promoting diversity in the field of computing.
• 📈 The speaker points out the overestimation of AI capabilities, particularly in facial recognition technology, which can lead to false positives and negatives with serious implications.
• 👥 The presentation emphasizes the need for systemic change within AI research and development to ensure that technologies are developed ethically and responsibly.
• 📚 The speaker discusses the importance of considering the 'back end problem' in AI, which involves the diversity of thought and backgrounds in AI research and development.
• 🤖 The presentation suggests that combining neural networks with probabilistic frameworks can lead to more robust and explainable AI systems.
• 🏛️ It calls for increased empathy in the field of computing and robotics, advocating for in-person sessions and active discussion to foster a sense of community and understanding.
• ✊ The speaker concludes by encouraging the audience to vote, emphasizing the importance of civic participation in driving change and ensuring equal opportunities for all.

Q & A

What is the main theme of the presentation?

The main theme of the presentation is the impact of AI mistakes on Black lives and society at large, particularly in the context of law enforcement and facial recognition technologies.

What is the significance of the hashtag 'scholarstrike' mentioned in the script?

The hashtag 'scholarstrike' is significant as it represents a movement that aims to highlight and address issues of social justice, which the presenter suggests will have profound effects on the field of artificial intelligence.

What is the role of blackincomputing.org as mentioned in the script?

Blackincomputing.org is mentioned as a resource that the presenter helped write, which likely serves as a platform to address the underrepresentation and challenges faced by Black individuals in the field of computing.

How does the presenter connect the death of George Floyd to the field of computing?

The presenter connects the death of George Floyd to computing by discussing the systemic issues that exist within society, including in the field of computing, and how these issues can be addressed through technology and systemic changes within the field.

What is the case of Robert Williams mentioned in the script, and why is it significant?

The case of Robert Williams is significant because it illustrates the potential for facial recognition technology to falsely arrest individuals based on incorrect identification, highlighting the need for more accurate and fair AI systems in law enforcement.

What is the 'back end problem' in AI research referred to in the script?

The 'back end problem' in AI research refers to the systemic issues within the field, such as the lack of diversity among researchers and developers, which can lead to biased and unfair AI systems.

Why does the presenter argue that the capabilities of neural networks have been overestimated?

The presenter argues that the capabilities of neural networks have been overestimated because they can still produce significant errors, such as false positives and negatives, which can have serious consequences in real-world applications like law enforcement.

What is the proposed solution to the issues with AI and facial recognition as discussed in the script?

The proposed solution discussed in the script involves embedding neural networks into a more robust, explainable, and introspectible probabilistic framework to improve the quality of AI decisions and inferences.

How does the presenter suggest improving diversity in AI research and development?

The presenter suggests improving diversity in AI research and development by addressing double standards, investing in diverse individuals, and ensuring equitable opportunities and outcomes, as mandated by laws like the Civil Rights Act of 1964 and Title IX.

What is the importance of empathy in computing and robotics according to the presenter?

According to the presenter, empathy in computing and robotics is crucial for creating fair and inclusive technologies. It involves understanding the impact of technology on all individuals, especially those who have been marginalized, and ensuring that everyone's voice is heard and considered.