What Role Does Instinct Have in an Increasingly AI-Focused World?
“Can machines think?”
Alan Turing first asked this question in 1950, and we’ve been asking it ourselves ever since. Artificial intelligence is already integrated into our every day; it suggests replies to our emails, recommends products as we shop, shows the fastest route to our destination. Last year an AI-created piece of art was auctioned at Christie’s for upwards of $400,000—and this is just about where people start to veer from fascination into fear.
When we ask, “Can a machine think?” we mean, “can a machine think like me?” and if it can think like a human then what separates us? What’s at stake?
Thinking thoughtfully about the issues tied to AI (ethics, job security, bias, etc.) is important—especially as it gains momentum. But, it’s also important to highlight how AI can help us tap into our potential by embracing the elements that make us human.
Particularly, instinct and creativity.
Instinct as a Guide to Learning
Part of our blueprint in building AB Tasty was giving people the autonomy and bandwidth to follow their instincts when it comes to website optimization. We tend to draw a line between data and intuition. One is rational, rooted in fact, the other based on incomplete information. So, it may be interesting that two founders of a tech company—both from web analytics backgrounds—were focused on empowering people to pursue gut feelings.
Psychologists at Tel Aviv University School of Psychological Studies conducted a study to test decision making based on instinct. Volunteers viewed two columns on a computer screen that each held a sequence of numbers, and were asked to determine which side had the highest average. Two to four pairs of numbers were quickly shown on the screen before being swapped out so participants didn’t have time to make calculations.
The results showed that participants had a high rate of accuracy during the exercise—which grew as more data was introduced. When analyzing six pairs of numbers, the accuracy rate among participants was at 65%. At 24 pairs, it grew to 90%.
The experiment showed that humans are highly adept at processing many pieces of information and intuitively assessing value. Put another way: humans can make decisions based on instinct that are pretty spot on.
Does this mean gut reactions are foolproof and enough to sustain a business strategy? Of course not. But they can serve as a springboard. At AB Tasty, we want people to tap into their instincts as a source of inspiration for testing, to validate their ideas with data or understand why they didn’t work—to let instincts guide learning.
Which is where AI comes into play.
How Artificial Intelligence Powers Faster, Better Experiments
Today, speed and accuracy are essential. At AB Tasty, we built our tool to be an accelerator for web optimization that improves over time, to streamline all the moving parts that go into digital campaigns. Once we achieved that, we asked ourselves: how can we make it even better?
At AB Tasty, we’re putting AI at the heart of our experimentation and personalization strategy. Currently, we use AI to:
- Dynamically allocate traffic to increase the speed and efficacy of experiments
- Match the best-converting product images to the right website visitors
One of the first ways we did this was with Dynamic Traffic Allocation. This algorithm modifies the quantity of traffic sent to each live test version to limit the “regret” of a test (or the loss of conversions that can occur when a part of your website’s traffic is sent to a “losing” variation).
Based on the data collected, website traffic is automatically redirected to more successful variations, to ensure the loss of conversions is limited during the experiment.
We then used AI-powered automation to drastically increase the speed at which a person or team can run conclusive A/B tests on images. Our Image Matchmaker widget automatically runs a series of tests to determine which visual on a product page performs the best, based on user-selected segmentation. Once the algorithm detects the uncontested winning image, it keeps that visual on the page for all incoming traffic.
All experimentation comes with a certain degree of uncertainty—a risk. We use artificial intelligence in our platform to make people more resilient to risk-taking by making sure each test has the most efficient outcome.
Innovation: The Intersection of Creativity and Data
The CEO of the design firm IDEO, Tim Brown, recently wrote about the competitive advantages humans can have in the time of automation, saying, “As the pace of innovation speeds up, disrupting entire industries, we must hold dear what is uniquely ours: The ability to understand human behavior and creatively solve for human needs.”
In our industry, solving for human need is a matter of understanding how humans interact with digital platforms. Data can tell us at what stage someone leaves a website or what element on a web page catches the most attention. But it can’t tell us that people are frustrated by a complicated checkout process, or more interested in a video than a static image. These are components that we have to infer with empathy, another trait of being human.
At AB Tasty, we’ve been using AI to optimize user experiences and the entire customer journey. However, the best uses of our platform will always rely on the inimitable inputs of human creativity and instinct.