We all know how impactful AI is today. However, in order to ensure that these technologies are developed and used in a responsible and equitable manner, there are ethical implications we must consider.
By popular demand, due to a successful speaker session about the importance of AI Ethics, Emmanuel Klu, Software Engineer at Google, is back to answer your questions about the topic.
Key Takeaways
Emmanuel Klu is a software engineer at Google working in the realm of responsible AI. He sat down with our Ciera Blanks to answer questions about and the current and future states of Artificial Intelligence (AI).
- How do you use AI responsibility? Have a familiarity with the different concepts of fairness in society – culturally, legally, philosophically, and morally. Also, consider privacy (making sure it doesn’t reveal personal information), robustness (it’s not making things up), transparency, and accountability (understanding why a model did something and the reasoning behind it).
- How do you avoid unintentional harm with AI? If you don’t intervene in any way, it’s just a reflection of the data it is trained on, so assume there is bias. It takes action and intervention to avoid perpetuating existing bias in the data.
- What is the best way to follow what’s going on in AI? Follow the research. By the time information gets to mainstream news, it has been shaped. Research journals and conferences are where cutting-edge info is being discussed. Follow academic research resources. Archive is a free source of research papers.
- What about AI and automation taking away jobs? The scientific perspective is to explore how to make everything more efficient; we need to pause and consider the long-term effects of automation. You may solve a short-term problem, but what does that look like in 5 years – is it still good? We should look beyond whether we can automate to the dynamics of how automation will affect humans over time and whether that helps society.
- What courses and concepts can you pursue to become more proficient in AI and language models? AI is very broad, so a foundational understanding of machine learning is very important. To narrow your focus, think about AI in two dimensions. First is the modality, text, speech, vision, and tabular data. The other dimension is machine learning tasks, classification, prediction, recommendations, or generative. Generate a combination of the two dimensions to find your interest (generative speech, text classification) so you can focus your learning.
The views and opinions expressed in this video are those of the presenter and do not necessarily represent the view of edX or its parent company 2U, Inc.
