Our Planet Powered by AI

7An Ecosystem and Investment Approach to AI Adoption

How can we make AI practical and accessible? This chapter aims to answer that question.

When something is described as practical and accessible, it often means that there are few barriers to entry and the process is not overly complex. For AI to become practical and accessible, we need to focus on ways of making it easier to use, reducing costs associated with developing and using AI technology, and creating an equitable environment that encourages diverse perspectives.

One proposed approach to reduce the complexity of AI development and make it more accessible is through no‐code or low‐code platforms. These platforms allow developers to create applications quickly, with minimal coding. This approach has been adopted by many tech giants, including Amazon with its Honeycode platform. This type of technology opens up opportunities for companies that may not have access to traditional development resources or expertise.

Another approach is to reduce the amount of data and computing power needed to develop an AI application. Founded by Rachel Thomas and Jeremy Howard, Fast.ai has developed a training technique that can allow applications to be built with just 30 data points, without needing expensive computing power. Moreover, the cost of access to a cloud‐based GPU has been reduced to just 45 cents per hour. This type of technology helps reduce the cost and complexity of AI development, allowing developers more access and flexibility ...