We need to broaden the efforts to expand ML/AI literacy around the globe.
Machine learning and Artificial Intelligence (AI) are fields of data science that have generated a lot of excitement about the potential applications that they enable. These applications span a wide spectrum of fields. In the field of games/strategy applications AlphaGo by Deepmind was able to defeat the human world champion in the Go game. In Text comprehension and generation, GPT-3 models can generate entire coherent articles. In voice and speech recognition, Alexa now supports more than 40 million users in the US alone. And recently, in the domain of creativity, AI DALL-E can generate beautiful original images based on a verbal description. And the list goes on and on.
For every famous innovation, there are many more machine learning and AI innovations hiding in plain sight. These are the ones we use every day without realizing. The applications that have already become mainstream and integrated with a lot of devices that we use day to day like, our phones and computers. Yet there is a huge lack of knowledge in our general population about how these applications work or how they are built. This lack of ML/AI literacy is even worse when it comes to resources in a language other than English, which most of the knowledge and technical applications are generated in. I am among the 94% of humans for whom English is not their first language. When I searched for resources that explain machine learning and AI in my native Arabic, there were close to none.
The absence of this accessible information is not only an ethical issue about diversity, representation and fairness in these fields which are shaping so much of our collective future. This lack of ML/AI literacy also has practical implications because people are using tools they do not understand. .
You might think that the lack of resources in other languages is due to a lack of interest. But that is not the case. A little over a year ago, I gave a talk in Arabic at The Palestine Digital Activism Forum 2021 about Machine Learning & AI applications, how it is built and the general ethical challenges that come with it. Given entirely in Arabic, the talk had more than 33,000 views in its first day. This highlighted the importance of the topic and the importance of talking about it, but this also has highlighted the challenge of ML/AI literacy even more for me.
We need to start a conversation about ML/AI literacy. Everyone should be familiar withI the basic concepts, their applications and uses and how they affect each and every one of us. ML/AI literacy should be included as apart of basic education in high schools in every country and every language. This includes both information that speaks to the general population and also specialized materials that can help people to understand complex topics. Organizations that already have strong ML/AI literacy curriculums should make the content available for citizen translators to access and make accessible to their communities. I personally would be eager to collaborate on a translation to Arabic, because I saw firsthand the hunger in my community for this information and they deserve to be part of building our collective future.