How to Create an Ethical AI System
So far on “Are You A Robot?” we’ve been discussing the necessity of having an ethical AI system. But how exactly do you build that? Shea Brown joins us to discuss how to do so and why auditing algorithms is necessary.
“Most people focus on the little piece which is machine learning part. But, you have to look at the whole chain to figure out where the opportunities are for people to get harmed.”
In this episode, Shea shares the important work himself and the team at BABL AI are doing to audit machine learning algorithms. The audit looks at the whole machine learning process, from beginning to end. Some auditors might just look at the machine learning part. But, it is necessary to look at how the data used to train the model was collected, to who and how they’re using the output.
Although audits are important in making AI ethical, he calls for more action on regulation that is strictly enforced. Although regulations might lead to less agile innovation, he explains that the cost benefit of being on the right side of AI is worth it. Once it becomes an industry standard, regulation will be less of an issue.
Shea believes that whilst we are waiting for regulation, companies need to be transparent as to how they are using machine learning and AI. He explains that many companies already are trying show their transparency to their consumers. For example, in the e-commerce sphere, whilst you are browsing, you might see a box pop up explaining why you are seeing that specific advert. However, Shea wants more. He thinks that there should be an option to step out of the tracking algorithm.
What other factors should be taken into account in AI auditing? Join our Slack community and let us know what you think!
What are your thoughts? Join our Slack channel and join the conversation!