Half-Human, Half-Machine, All Marketing: Managing Algorithmic Limitations
When AI-based algorithms were introduced to the marketing world they were an immense game changer: they upped the ante on how brands could target their existing audiences, and they sped up the process of understanding how to build new ones. As machine learning grew exponentially, however, it also lost much of its human touch.
Without a balance of human input and safety checks, algorithms can adopt human biases (including hate speech), and cannot adapt fast enough to real-time crises like COVID-19. The resources here discuss how marketers can best work with algorithms to ensure they reach desired targets with efficient, optimized results, and avoid communicating the wrong messages.
Check out the resources below.
Proactively Detecting and Reducing AI Bias. Pega, July 2020
Self-learning algorithms are designed and trained by humans, and humans have flaws. We're liable to pass on those flaws to the AI – intentionally or not – because of how we collect data, train models, and apply rules or logic when making decisions. We end up passing down algorithmic bias, and that becomes a huge liability – one that every business needs to be constantly aware of and proactively working to eliminate. Since the mistakes of bias can't be fixed after the fact, we need to be proactive in our prevention. Here are some actions that you can take to help identify and remove potential bias:
- Consider data quality as the utmost importance.
- Seek input from diverse and inclusive collaborators.
- Take an "always-on" approach to bias protection.
- Get ahead of future regulations. Know your and your customers' values and adjust bias to them.
A Guide to Understanding AI's Weirdness. Tech Talks, July 2020
These days, it can be very hard to determine where to draw the boundaries around artificial intelligence. What it can and can't do is often not very clear, as well as where its future is headed. In fact, there's a lot of confusion surrounding what AI really is. Marketing departments have a tendency to somehow fit AI in their messaging and rebrand old products as "AI and machine learning." Meanwhile, social media is filled with examples of AI systems making stupid (and sometimes offending) mistakes. This article summarizes a new book that discusses what AI – specifically, deep learning – is and what it isn't, and how we can make the most out of it without running into pitfalls. One humorous example is shown below, among others that pose a greater threat to brand safety:
Marketing and Technology Data Is Racist and Biased: Here's How Marketers Can Fix It. Search Engine Journal, June 2020
With the resurgence of the Black Lives Matter movement on social media, many marketers are taking new notice of issues that BIPOC (an acronym that stands for Black, Indigenous, and People of Color) and allies have been talking about for years. The most notable of these is how data – which we traditionally deem as unbiased and "just the numbers" – is in fact very influenced by the biases of the engineers, marketers, developers, and data scientists who are programming, inputting, and manipulating that data. This piece discusses what implicit bias is, how it affects marketing data and technology, why it's important to recognize it in our SEO and marketing, and what marketers can do.
Data Science: Why Humans Are Just as Important as Math. The Customer, July 2020
While today's data science tools can sift through mounds of data to unearth patterns at levels of scale and speed that humans alone could never achieve, our models remain inadequate in fully understanding data and its applications, especially when the data becomes messy in reflecting fickle human behaviors. Data science is a craft that relies on human intuition and creativity to understand multi-faceted problem spaces. Without human oversight, it operates on an incomplete picture, for which the implications have never been clearer in the present COVID-19 age, as our algorithms struggle to grasp the reality that human behaviors don't follow mathematics.
Content Strategy: Writing for People or Algorithms? Insightly, July 2020
Writers are inspired to tell interesting stories, pass knowledge, and/or entertain their readers. But content saturation complicates things, and when writing for business especially, they have to adapt their work (enter SEO content) to actually reach their target customers and broader audience. So how do you find a balance between writing content that solves problems and provides value to your audience, while at the same time appeasing Google's changing algorithm? This piece offers tips for finding a balance between writing content for SEO/readers and leveraging CRM data.
Stay on top of trends with Marketing Futures Pulse issues, which explore how new technologies, current events and trends, and innovations will affect marketers and consumers alike.
Connect with our Ask the Expert team in real time for customized answers to your specific marketing challenges.
Submit a request to Ask the Expert here.
Josch Chodakowsky is a senior manager of research and innovation at ANA.