At the BRITE '22 conference, Renée Richardson Gosline (Senior Scholar, MIT Sloan School of Management) spoke about the massive commitment companies are making to invest in artificial intelligence (AI), highlighting an IDC analysis that more than $120 billion will be spent by 2025, a doubling from 2021 investment.
When it comes to consumer experience, there is a focus on using AI to reduce customer pain points by removing the friction from these experiences through eliminating or automating processes and optimizing personalized experiences. Importantly, though, Gosline notes that "AI is often assumed to be 'neutral' and it is absolutely not." She highlights numerous documented cases where machines have incorporated biases -- on image identification, gender determination, health decisions, etc. -- that existing in the data used to train them.
"Instead [of thinking] of friction as tantamount to a pain point and something to be eliminated," states Gosline, "I like to think about it in the way that physics thinks about friction. In physics, friction is neither good nor bad. It just is. In fact, you need friction in order to accelerate... that is good friction.... This is why I want us to think about good friction as we think about the use of AI in customer experiences."
To wit, she presents four principles companies should apply to eliminate bad friction and add good friction:
- Ask if this is a decision that AI can be trusted to make
- Use AI with customer data to advocate for greater shared value
- Monitor customer journeys to uncover where friction is helpful
- Use AI to create human communities among your customers
Within an organization, she also notes that the insights of AI are likely to be most powerful when they aren't used solely for automaton -- e.g. using natural language processing to churn out simple ad copy -- but when a moment of friction is introduced by flowing the insights through employees who have human understanding and creativity to advance or critique the efforts of the machine.
"A point of differentiation for brands going forward is going to be people feeling comfortable with the way their data is being used and they way these models are being applied," notes Gosline in conclusion, "And the firms that have the discipline to resist the siren song of extracting as much data as possible under the guise of frictionless are going to be best positioned to built trust and loyalty in the long-term."