8.
Prototype
8.1 Which AI tool(s) will you use to build your prototype?
PyTorch: Deep learning frameworks for building and training neural networks.
8.2 Which AI tool(s) will you use to build your solution?
8.3 What decisions or outputs will your tool generate and what further action needs to
be taken after a decision is made?
18
9. Testing
9.1 Who are the users who tested the prototype?
Cultural Anthropologists and Sociologists:
Researchers studying cultural diversity and human migration patterns.
Academics interested in understanding the distribution of ethnic groups across
continents
9.2 List your observations of your users as they tested your solution.
They appreciated that the AI seemed to avoid stewreotypes and biases in its
classifications
They found the interface easy to use, and the results being presented in a clear and
understandable format.
9.3 Complete the user feedback grid
What works What needs to change
Speed: The system provides Data Bias:Make sure the Ai
results quickly, without does not promote or provide
excessive processing time. results which are biased
User-Friendliness: The interface towards certain ethnic groups
wis easy to navigate and or demographics
understand, even for users who Explainability: It would be
are not tech-savvy. helpful to understand how the
Privacy Protection: "I felt AI arrives at its
confident that my personal data recommendations so that I can
was handled securely and trust its decisions.
responsibly." Data Diversity: "Consider
expanding the dataset to
include a wider range of
ethnicities and individuals from
different socioeconomic
backgrounds."
19
Questions? Ideas
How would you prevent the
potential misuse of AI in
healthcare, such as for
discriminatory purposes.
How does the AI account for
individual differences in
preferences and needs?
How often are the
recommendations updated
based on new data?
9.4 Refining the prototype: Based on user testing, what needs to be acted on now so
that the prototype can be used?
Privacy Protection: Ensure that user data is handled securely and ethically.
Transparency: Be transparent about the AI's decision-making process and
limitations.
Infrastructure: Ensure the AI can be deployed on appropriate hardware or cloud
infrastructure.
Data Quality: Ensure the AI is trained on high-quality, diverse, and representative
data to improve accuracy.
9.5 What improvements can be made later?
Advanced Algorithms: Explore more sophisticated machine learning algorithms to
improve accuracy and reduce errors.
User Preferences: Incorporate user preferences and feedback to tailor recommendations
more accurately.
Habit Formation: Analyze user behavior patterns to predict future needs and preferences.
Natural Language Processing: Improve the AI's ability to understand and respond to
natural language queries.
20
10. Team collaboration
10.1 How did you actively work with others in your team and with stakeholders?
Data Analysis and Preparation: I would work closely with data scientists to analyze and
prepare datasets, ensuring they are representative, unbiased, and suitable for training
the model.
Algorithm Development: I would assist in developing and refining machine learning
algorithms, suggesting improvements based on my understanding of the data and the
project's objectives.
Model Testing and Evaluation: I would help evaluate the model's performance, identify
potential biases, and suggest ways to improve its accuracy and fairness.
Collaboration with Stakeholders:
Understanding Requirements: I would engage with stakeholders to understand their
specific needs and goals for the ethnicity detector, ensuring that the model aligns with
their expectations.
Addressing Ethical Concerns: I would help identify and address potential ethical issues
related to the project, such as bias, privacy, and discrimination.
Providing Insights: I could provide insights into the model's decision-making process,
helping stakeholders understand how the model works and why it makes certain
predictions.
Key Collaboration Techniques:
Regular Communication: Maintaining open and frequent communication with team
members and stakeholders is essential for ensuring that everyone is aligned and
informed.
Feedback and Iteration: Incorporating feedback from team members and stakeholders
allows for continuous improvement and refinement of the model.
Transparency and Accountability: Being transparent about the model's limitations and
potential biases helps build trust and accountability.
21
11. Individual learning reflection
11.1. Team Reflections
A good way to identify what you have learned is to ask yourself what surprised you during the
project. List the things that surprised you and any other thoughts you might have on issues in
your local community.
Team member name:Aayush
An information researcher in this project would gather, process and evaluate datasets
pertaining to the historical, cultural and genetic events that have determined the
physical characteristics of different races. The main aspects of this position would be
studying the rationale behind the existence of physical characteristics and the extent to
which they can be used for classifying people of different ethnicities into certain
groups.
Team member name: Aditya
As a data scientist, selecting the right data from the vast library of information
determines the success of the project. This process must be unbiased and adhere to
ethical AI rules, with privacy and security being the priority to prevent data leaks.
The project's effectiveness also relies on a strong team, good leadership, and clear
communication to achieve maximum efficiency.A good project will have good
foundation of data and be secure while boasting great efficiency.
Team member name: Kanishk
A video producer is responsible for managing the entire video production process,
from start to [Link] producers should have excellent communication,
organization, and time-management skills, as well as a keen eye for detail. They may
also need to be proficient in cinematography, editing, sound design, animation, and
marketing.
22
Team member name: Pratik kothari
As a tester in the ethnicity detection AI project, I learned the critical importance of
comprehensive test planning and the balance between automated and manual testing. I
focused on evaluating data quality to mitigate bias and developed specific test cases to
ensure fairness across demographic groups. Collaborating with data scientists and
stakeholders deepened my understanding of ethical implications and reinforced the
need for responsible AI practices. This experience has significantly enhanced my
testing skills and adaptability in a rapidly evolving field.
Team member name: Raghavan
Identifying a person's continental background based solely on facial features is
challenging due to variations within and between ethnic groups. AI systems trained on
biased data can perpetuate stereotypes. While this technology has potential
applications, ethical concerns like misuse and discrimination must be addressed. Deep
learning models and multimodal approaches could improve accuracy, but careful
consideration of ethical implications is crucial.
Team member name: Sai haresh
Developing an AI ethnicity detector has highlighted the importance of addressing bias,
ensuring data quality and diversity, selecting appropriate models, and considering real-
world applications. Mitigating biases in the data and model is crucial to prevent
perpetuating harmful stereotypes. Collecting diverse datasets and addressing data
imbalances is essential for accurate and fair results. Choosing the right AI model and
evaluating its performance rigorously is key to ensuring reliability and interpretability.
Considering the social and cultural context of the application is vital to avoid
overreliance on the technology and ensure responsible use.
Team member name: Sakthi Shabarieshvar
Leading the ethnicity detection AI project enhanced my skills in project management,
team dynamics, and ethical considerations. I learned the importance of clear objectives,
effective resource allocation, and open communication, which fostered collaboration
and constructive conflict resolution. Gaining insights into algorithms and the
significance of data quality underscored the need for diverse datasets to mitigate bias.
Additionally, engaging with stakeholders highlighted the ethical implications of our
work. This experience has equipped me to balance innovation with responsibility in
future projects.
23
