After-Event Recap

The ideathon, held on 7 th September 2024 at Kerala University of Digital Sciences, Innovation and Technology, focused on the critical issue of making Artificial Intelligence (AI) more responsible and inclusive. This event attracted participants from diverse backgrounds, including students, researchers, and professionals, all of whom brought forward innovative ideas and practical solutions to address the ethical challenges surrounding AI technologies.

Event Overview

The ideathon was judged by two esteemed professionals:

•  Aswin V S, Assistant Professor at Kerala University of Digital Sciences, Innovation and Technology

•  Jibu Elias, Country Lead for India at the Responsible Computing Challenge, Mozilla Foundation, an organization known for its dedication to promoting ethical computing and responsible use of technology.

Objective

The primary goal of the ideathon was to explore how AI can be developed and deployed in ways that ensure fairness, transparency, and accessibility for all individuals, regardless of their socio-economic background, race, or gender. It emphasized the importance of ethical frameworks in AI development, the need for diverse representation in AI training datasets, and the mitigation of algorithmic biases.

Key Themes Discussed

1. Bias and Fairness in AI: Participants discussed how biases can be inadvertently embedded into AI models through biased training data and algorithms. Solutions proposed included:

• Developing diverse datasets that represent all demographic groups.

• Balanced Content Curation with Filter Bubble Management.

2. AI Accessibility: There was a strong emphasis on making AI technologies accessible to underserved communities. Ideas included:

• Designing AI systems that are adaptable to different languages and regional needs.

• Ensuring AI solutions address issues relevant to marginalized groups.

3. Transparency, Privacy and Accountability: Several teams proposed strategies to enhance the transparency of AI systems. Ideas included:

• Creating mechanisms for users to understand how AI systems make decisions.

• Developing policies that require companies to disclose AI's decision-making processes.

• Digital Watermarking and Content Authentication Networks

4. Collaboration Across Stakeholders: The role of government, industry, and academia in ensuring AI is developed responsibly was a major point of discussion. Participants proposed:

• Collaboration between public and private sectors to create ethical AI guidelines.

• Involving underrepresented communities in the AI development process.

Winning Ideas

The winning team presented unique solutions regarding the genuinity of contents from Gen AI that aligned with the principles of responsible and inclusive AI. These solutions addressed:

LLM’s should consider the feedback of majority of the public which helps in increasing the genuinity of information.

LLM’s can be trained to accept their uncertainty about their knowledge rather than providing false information.

A platform can be made where the LLM can learn from the public.

Conclusion

The ideathon concluded with insightful feedback from the judges, Aswin V S and Jibu Elias, who appreciated the participants creativity and commitment to responsible AI development. Both judges emphasized the importance of continuing discussions around responsible computing and inclusivity, encouraging participants to pursue their ideas and contribute to a more ethical AI future. This ideathon was a testament to the growing awareness of the ethical responsibilities in AI development and the need for inclusive, fair, and transparent AI systems. It showcased the potential of collaborative problem-solving in addressing some of the most pressing issues in AI today.