Sign up for our newsletter and get the latest big data news and analysis.

Implementing Ethical AI

In this special guest feature, R.J. Talyor, Founder and CEO of Pattern89, discusses implementing ethical AI into a company’s business strategy. Pattern89 is an artificial intelligence-based software company that optimizes paid social media advertising campaigns. Previously, R.J. was the driver behind ExactTarget’s mobile strategy, including the launch of SMS marketing in 2002 and spearheading mobile marketing during the launch of the iPhone. His leadership was also instrumental when Salesforce acquired ExactTarget for $2.5 billion in 2013, where he became the vice president of mobile products.

Last fall, an app hit the web to classify human beings with artificial intelligence. ImageNet Roulette is part of an art exhibit on the history of image recognition systems. The app showed the bias and flaws associated with categorizing human faces by machine, and outside of the context of the art exhibit, it was downright racist.

AI has the power to analyze billions of data points in the blink of an eye and translate it into actionable insights. It would take a human more than 100 lifetimes to process that much data. For marketers, tools such as natural language processing and computer vision can translate data into marketing components that are guaranteed to have the greatest return. As a result, marketers can streamline strategy and execute campaigns at the right time and place with the right copy and photo.

Herein lies the problem. Computer vision looks at images, and those images include humans. AI was created by humans and, within it, carries human biases. It measures what a human tells it to measure, aggregating a lifetime of knowledge based on a human directive. So, if that human directive is biased, the AI is biased and will learn more through that biased lens. Even if the AI is built with noble intent, humans can still develop this technology with objective and personal opinions that make it deeply flawed.

As more industries implement AI into their processes, marketers specifically must use AI in a bias-free manner to measure their ads based on ethical interpretations of images. This can be done with the right tools and the right humans behind them.

Implementing Guidelines

In April 2019, the High-Level Expert Group on Artificial Intelligence presented “Ethics Guidelines for Trustworthy Artificial Intelligence.” Per this report, there are seven key requirements that AI systems must meet in order to be deemed “trustworthy,” including human agency and oversight, transparency, diversity and more. While these guidelines are a great place to start when implementing AI, one continental legislative body cannot make guidelines for the whole world. AI is built in different nations with different contending biases.

Using established guidelines as a jumping-off point, a company should also compile several advised guidelines into its own ethical principles to reflect the makeup of its team and core values. This means building or working with a team that represents a country in terms of gender, age and ethnicity. In the U.S., that team would ideally be 50% female and 27% people of color. Committing to ethics and diversity allows the humans behind the AI to ask necessary questions, and therefore, think of the best solutions possible.

Adhering to Best Practices

AI adoption is still young, and many companies lack a strategic focus on its integration. Unfortunately, there’s not a one-size-fits all approach, so technological best practices must be incorporated into any new AI tools. This includes understanding how AI learns, how it prescribes tags to images and words, and how data come together to serve users recommendations.

In addition, teams must also be mindful of eliminating biases in race, religion and other dimensions. Eliminating these human biases and prejudices means AI works objectively to make the best determinations for its users.

Having Open Conversations

Transparency is a key ethical requirement for AI. No matter the industry, companies using AI need to be able to explain to users how their AI works, how the business works and how it affects outcomes of the technology. Companies that are unwilling to be transparent about their AI should raise a red flag – are they hiding that their AI may not be true AI, but machine learning with regular human inputs? Do they have something in their algorithms that is unethical they want to keep confidential?

Transparency shouldn’t just be kept to the tech, either. Informing those within your organization about the implications of AI, how it works and how it complements their jobs is key to improving the success of AI.

AI is an emerging technology taking the world by storm. In marketing, it transforms how businesses communicate with customers and influences purchasing habits and the ads they see. Organizations are responsible for ensuring AI is implemented in an ethical manner. Infusing AI into a marketing strategy, or into overall business operations, is more than just a flashy new technology – it can have a profound impact. Therefore, it is imperative to ensure ethics and bias are being addressed from the beginning to achieve the best results.

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: