Abstract
- Decision bias: Relying solely on data to make decisions often affirms bias rather than optimizing results. Look for diverse data sources to challenge your preconceptions.
- Consumer vs. producer: The impact of data varies widely between consumers and producers, influencing the decision-making process differently based on perspective.
- The role of AI presents challenges such as: Rebecca Haddix urges decision makers to ask tough questions of AI and argues for its role in not just confirming business strategy, but challenging it..
A few months ago, I was doing research for a blog I was writing about data-driven decision making. Regular readers of CMSWire know that there is a lot of information out there on the subject, from pseudoscientific studies on the effectiveness of human intuition to eerie predictions that humanity will soon cede all control to his AI. Masu.
But what is clear is that when it comes to using data in decision-making, analysis is too often used to affirm preconceived notions rather than to optimize outcomes. Let's look at some issues related to decision bias.
Data abuse: Skepticism in statistics is healthy
The idea that data can be misleading and misused is not new. Mark Twain spoke of “lies, outright lies and statistics,” quoting British politician Benjamin Disraeli. Most people attribute this quote to Twain, which is not true, but it is also actually doubtful whether Dillaelli first said it. My point is that whether you get your facts from tabloids, chatbots, or analytical dashboards built into DXPs, skepticism is healthy.
Related article: Are your business decisions failing because of bias?
Data-driven decision making: Doing more harm than good?
While researching opinions on this topic, a Forbes article written by Rebecca Haddix caught my eye. I found his sincerity and authoritative tone refreshing. I also liked the title, “Your Data-Driven Decisions Are Probably Wrong.” Rebecca has been a technology contributor on big data for Forbes magazine for nearly a decade.
Her article, published in 2020, didn't mention generative AI at all, but the strategies and clear guidelines she proposed are relevant to our current debates about bias and data in decision making. It seemed so relevant to the discussion that I felt compelled to reach out. We wanted to know if her advice for decision-makers in 2024 has changed now that AI is in every conversation.
She was kind enough to arrange a video call with me. Below is a summary of some of her deep insights from that discussion. Consider what Rebecca has to say about things like data-based decision-making and decision-making bias.
Related article: 3 ways to reduce bias in customer survey data and achieve effective CX
The impact of big data, for better or worse — it's a matter of perspective
The vast amounts of data we collect can be good or bad, depending on your perspective as a consumer or producer. Read any study and you'll see that daily data creation is now measured in hundreds of terabytes per day, and the rate of collection is expanding exponentially.
When we asked Rebecca whether the vast amount of data we collect is good or bad for us, her answer was nuanced and surprising, and it differentiates consumer and producer decision-making. did. She replied: “I think it depends on who we become in that equation. Right? So, as consumers, we have a choice between producers and products created by companies. ”
Related article: Addressing bias in AI, Part 1: Recognizing bias
Consumers delegate choices and trust overrides optimal decisions
She went on to explain that consumers want to delegate to trusted sources, influencers or search engines. Consumers get excited about the promise of making the best decision, but it's often not what they really want or need.
“Search is based on many factors, such as marketing and SEO, and there are many factors that we don’t consider, which can lead to us not making the best purchasing decisions as consumers. That’s why it’s important to have an aggregated source that people trust. “As consumers, we've never made optimal decisions. More data means less analysis.
“But when it comes to us as producers, as companies, as creators of these products, I think the increase in data is really exciting, as long as we're intentional about how we process and analyze it. .
“The human brain as a consumer has already reached a point of saturation and can't take any more. So I'm really excited about the rise of big data. And that, whatever it is, The impact on companies that are producing new products and optimizing how we actually work is having the ability to act on a greater understanding of the real relationships between separate things. It's fun.
Related article: The importance of data literacy in business decision making
It is human responsibility to ask the right questions
Rebecca previously wrote: “Outputs are only actionable if the inputs are relevant.
“We can get answers. Based on training data, we don't necessarily get answers to the questions you ask. So it's our responsibility to ask the right questions and make sure that is to do it. [the AI] It is trained on a suitable dataset. ”
For example, if we were looking for information on a medical topic, she suggested: “From these dates by these authors who have been cited at least her X times, we can only look at her JAMA Journal of the American Medical Association and answer: 'Which treatment is most effective for this condition? Questions like “?” Next, let's verify it.
Related article: Overcoming AI bias in CX with Latimer
Challenge AI: Avoid bias and enhance your marketing strategy
“So when it comes to marketing technology, specifically, 'The campaigns we can run and the data we can collect only reflect visitors who reach the site through existing channels. We have But there may be optimal implementations and experiments that we don't have yet.”
She goes on to suggest that the entire process begins with a clear problem statement and four or five hypotheses about the best solution before asking the question. [to AI]And at that point, you say, “Instead of asking the AI to prove you right, let's dare the AI to prove you wrong.'' This helps avoid bias in decision-making. Helpful.
Decision Bias: It's human nature to want to be right
Rebecca emphasized the importance of “continually checking our intuition against our own biases, because when we have an idea, we want to be right, so when we look at the data ourselves, we want to be right.” The scientific method doesn't allow us to do that.''We keep looking for evidence that our hypothesis is wrong. In doing so, we can avoid this loop of circular bias where we all want to be right. ”
Rebecca added: “I like to say that our intuition is just as important as our data. [but] It's about asking the right questions and using data to disprove assumptions, rather than just leaving decisions to dashboards. ”
(Author insight: Rebecca Haddix is not a fan of analytics dashboards.)
Isaac Asimov was a visionary
Towards the end of the interview, Rebecca recited sci-fi author Isaac Asimov's three principles of robotics almost verbatim, which really stuck out to me. This reinforces my belief that AI should act like a guide dog to humans, focused on keeping us out of harm's way rather than accelerating human progress. It was born when I told her.
For reference, Asimov's three principles of robotics are: (1) Robots must not harm humans or inadvertently cause harm to humans. (2) Robots must follow orders given to them by humans, unless they conflict with the First Law. (3) Robots must protect their own existence as long as it does not conflict with the first or second law.
She talked a little about how these rules chain together and each builds on the previous one. This is a powerful concept, but it's worth noting that in order to respond to our invitation to become faster and more aware, we need to put up guardrails to protect and enforce restraint. That means they must be trained to identify and protect against hazards.
final thoughts
It was enlightening to hear thought leaders already building solid ideas over the past few years. The new information did not change the previous advice, but the deeper insight enhanced it.
Much like famous thinkers like Isaac Asimov and Rebecca Haddix. I look forward to her future articles and further discussions on data-driven decision making, decision bias, and other topics.
Find out how to join our contributor community.
