By Steve Southin, Co-CEO, PAVE
The COVID-19 pandemic has forced many retail businesses to either shut down or embrace automation, AI, and digital retail solutions in order to continue to generate revenue. Almost immediately, artificial intelligence has gone from a niche concept to a one that many automotive dealers and businesses need to strongly consider even as many of us begin the re-opening process. While some people believe it could unlock a new golden age for humanity, some, including the late Stephen Hawking think that it could plot out humanity’s doom. Are the robots really here to take jobs, or does AI need human intelligence to exist and be nuanced enough to really benefit retail?
Aside from the data scientists and programmers who work in the automation field, think about all the people who perform tasks that AI’s aren’t equipped to handle, thereby helping to improve them, and vice versa.
To compare the combination of man and machine to the two parts of a brain, like Microsoft’s Dr. Hsiao-Wuen Hon, consider machines as the left side of the brain (the side responsible for logic and rationality), with humans being the right side (the side responsible for creativity and judgment and decisiveness). Together, they can form an ideal form of intelligence, which Dr. Hsiao-Wuen Hon expresses in the following equation: Artificial Intelligence (AI) + human intelligence (HI) = intelligence, or augmented intelligence.
Amazon’s Mechanical Turk is often described as “artificial intelligence” even though it uses a small army of workers to carry out micro-tasks like matching information in photographs – something machines aren’t usually very good at. This is changing as machines learn more over time. Mechanical Turk workers end up spending more time checking the work of algorithms, as opposed to filling in ability gaps for them.
Another example of how AI solutions need Human Intelligence first can be found in the automotive vehicle inspection process. Over the course of 15 years, in North America, there have been over 22,000 body styles of vehicles produced. Each style can have up to 10 unique colors, creating hundreds of thousands of possible vehicle combinations. There are also over 1,000 combinations of types of damages and severity that AI would need to learn, detect, and process. It would be almost impossible for AI to consistently detect the same damage and severity on each different body style and color, as every different variation affects how this damage can appear for AI vision. The AI would need to learn over 2,000,000 unique variations and combinations AND be able to learn and detect new variations as they come into existence.
In order to avoid this problem, ideal AI solutions need to include Human Intelligence and cannot rely on AI to learn an impossible amount of data and execute nuanced functions that can take their human counterparts a short period of time to process.
While concerns about jobs being “lost to machines” may have some validity, new roles are popping up for those within the Artificial Intelligence arena because AI needs a human touch in order to benefit retail and provide the kind of user experience that retains customers and keeps businesses growing.
About the author
Steve and his 25+ years of automotive retail and wholesale experience deliver in-depth domain knowledge that was essential in his focus as PAVE’s creator and product architect. Steve also has 15+ years of technical and startup expertise that he gained as an Autotech entrepreneur, with his second recent successful exit being the Bumper App, which he brought to market in 2012 and successfully exited through an acquisition by Vicimus Inc. in 2017