In this episode, Lauren Hawker Zafer and Oriana Medlicott talk about Enterprise Modernization - AI Ethics; a space reserved for academics and specialists until very recently. Nowadays, however, data & ethics are relevant topics for IT companies around the globe. Not only because failing to adhere to ethics can pose tremendous reputational, regulatory and legal risks but also because it could lead to inefficient product development or resources management. With all of this in mind: what are the main challenges and implications of AI and ethics?
Bridging Cultural Differences in Ethics
What is considered ethical in AI always depends on the cultural background and contextualization of those looking at it. For example, Europe and the US, both have a similar understanding of ethics, while countries like e.g. China differ slightly in their ethical alignments and interpretations. For international companies with a worldwide presence, like e.g. Fujitsu, it is important that these differences can be bridged and that company-wide AI ethical standards that reflect the cultural nuances of all countries are established.
Bias Training and Team Diversity
Bias in algorithms and data sets is one of the most relevant issues when discussing ethics and AI. Bias training and the conscious choice to assemble diverse teams who compile and train AI data sets might be an option to solve bias-related issues. While the problem of biased data would be difficult to eradicate completely, the aforementioned measures can considerably improve the quality of data.
Will Artificial Intelligence Surpass Human Intelligence?
The question of whether humans are externalizing their intelligence in training models that are smarter and faster than them is an important consideration in AI ethics and one that Oriana and Lauren address in this episode of the podcast. With humans training AI models, there is a debatable fundament of the limitations drawn in alignment with those training the models and computational capabilities. After all, many human qualities surpass computational capabilities, such as the holistic contextualization of information or the application of an emotional layer.
Learn more about AI ethics by listening to the podcast!
About the Speaker
Oriana Medlicott is a senior researcher in the technology strategy unit at Fujitsu and a co-host of the Let’s Chat Ethics podcast. In her position at Fujitsu, she contributes to shaping the company’s technology strategy through technology market analysis and research, finding start-ups to partner with, and organizing events and external partnerships to promote research and development in areas such as AI ethics, data and security and many more.