Follow Us on Google
A Product Advisor for startups and entrepreneurs and leader of a high-performing team of product managers and designers, Kingsley Ukeje has revealed that openness, inclusivity and accountability are the key components of designing ethical and user-friendly AI system.
In this view, Ukeje who the Head of Product & Design, Eze Wholesale on Human-Centered AI, spoke extensively on designing ethical and user-friendly AI systems, the challenges, and the future of AI in terms of ethical design among others
Can you start by telling us a bit about your background and your journey in the fields of design and product development?
I shape the product vision and strategy for the world’s foremost B2B wholesale marketplace for smart phones, laptops, tablets, and other electronics. I have over eight years of experience in product management and design, and I am also a Product Advisor for startups and entrepreneurs. I lead a high-performing team of product managers and designers to collaborate with cross-functional departments to deliver products that drive growth, customer satisfaction, and market leadership. I leverage user research, Figma, and design systems to create user-centric and innovative solutions that solve complex problems and add value to our customers and partners. I am passionate about building products that make a positive impact on the world and empower people to access quality and affordable electronics. This journey has given me invaluable insights into designing, building, deploying, and managing digital products.
With your extensive experience, what do you think are the key components of designing ethical and user-friendly AI systems?
A multidisciplinary approach that gives users wants and rights priority is necessary for designing AI systems that are both ethical and easy to use. First of all, openness is essential. Users need to know how their data is used and how AI systems make choices. To guarantee that AI systems serve a variety of user groups and do not reinforce prejudices, inclusivity is also crucial. Thirdly, the system needs to be made accountable, which means there needs to be a defined process in place for handling any problems that may come up. Lastly, the development process should be guided by user-centric design concepts to guarantee that AI systems are accessible and understandable, and improve the user experience.
Can you share some examples of projects where you’ve successfully integrated these principles?
Certainly, one noteworthy initiative is Carrot Credit, where we made sure there was transparency by explaining to our users how their data is used for fraud detection and giving them access to our AI systems’ decision-making process. By creating algorithms that take into account a variety of user behaviours and patterns, inclusivity was given top priority and the likelihood of biased results was decreased. By putting in place a strong feedback system that allowed customers to voice any problems or complaints, which our team quickly resolved, we created accountability. Ultimately, by making sure the AI tools were simple to use and straightforward, our user-centric design approaches improved both the overall user experience and platform confidence.
What challenges do you typically encounter when designing AI systems, and how do you overcome them?
Managing biases in artificial intelligence is one of the major obstacles. When artificial intelligence (AI) systems learn from biased data, they may reinforce such biases. We employ methods to reduce bias and carry out extensive audits of the data we use to get around this. Maintaining user trust presents another difficulty. Users may be dubious about AI’s judgements because it is sometimes viewed as a “black box.” To tackle this, we prioritise openness and user education, assisting people in comprehending the AI’s functioning and streamlining the decision-making process. Finally, while it can be difficult to keep up with the quick speed at which AI is developing, we can stay ahead of the curve by working with industry experts and engaging in ongoing learning.
How do you see the future of AI in terms of ethical design and user experience?
I think that improving user experience and ethical design will become more and more important in AI in the future. Transparency, accountability, and diversity will be in increased demand as AI systems become more ingrained in our daily lives. To construct AI systems that are not only strong but also fair and trustworthy, designers and engineers will need to collaborate even more closely with ethicists, policymakers, and end users.
Furthermore, I anticipate that as AI technology develops, user interfaces will become increasingly smooth and intuitive, increasing the accessibility and usefulness of AI tools for a wider range of users. To make sure that these systems serve humanity as best they can, it will be crucial to maintain the human aspect at the core of AI development.
How do you balance the technical capabilities of AI with the ethical considerations that come into play, especially when dealing with sensitive data?
Although difficult, striking a balance between technical prowess and moral issues is essential. It begins with a thorough comprehension of the potential hazards and ethical ramifications of deploying AI. We have strong data privacy and security procedures in place when working with sensitive data, making sure we abide by laws like the GDPR. To safeguard user data, we also use methods like federated learning and differential privacy. To make sure our AI systems adhere to moral principles while utilising their technological prowess to benefit users, we also regularly carry out ethics evaluations and interact with a wide range of stakeholders, such as ethicists, legal professionals, and end users.
That’s excellent advice. Can you elaborate on any specific methodologies or frameworks you use to ensure your AI designs remain user-centric and ethical?
Indeed. Human-Centered Design (HCD) is one framework that I use a lot. There are three primary stages to this process: conception, ideation, and execution. Through surveys, observations, and interviews, we obtain comprehensive insights into the demands and challenges of users throughout the inspiration phase. We brainstorm and prototype solutions during the ideation stage, continuously seeking user feedback to improve our concepts. During the implementation phase, the solution is developed and deployed, and ongoing user testing is conducted to make sure it satisfies user expectations and ethical requirements. Using ethical checklists and guidelines—like the IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems—is another technique we employ to make sure our AI systems are developed with ethical standards in mind.
What role do you think regulation and policy will play in the future of ethical AI design?
Future ethical AI design will be greatly influenced by legislation and policy. Clear rules and regulations will be necessary to guarantee that AI is developed and utilised properly as it continues to advance and permeate more areas of our lives. Policies can help establish benchmarks for accountability, justice, and transparency by giving organisations a structure to work within. By rewarding ethical behaviour and discouraging unethical activity, they can also promote the advancement of ethical AI. To create legislation that strikes a balance between innovation and ethical considerations and guarantees that AI technologies serve society as a whole, collaboration between legislators, industry leaders, and the general public will be essential.
