By John Walubengo
Last month the Governor of Kisumu County issued a protest tweet and letter to one of the local airlines plying the Nairobi-Kisumu route. He told them they were overcharging, given that his one-way bill was double what he was used to paying.
In their response, the airline tweeted back a letter disclosing that they are a private company and are free to charge whatever is appropriate but mainly based on demand, peak travel days and time of booking.
While this is true, there is more that airlines and other mobility providers, such as online taxis, can do.
They can, for example, decide the price you pay based on static and dynamic data sets. Static data sets tend to be less sensitive, while dynamic data sets are more personal.
Dynamic data includes but is not limited to the device type you booked from- you book from a MacBook or iPhone, and you pay more than those booking flights from their Windows or Android devices.
The geo-location from which you make your booking can also affect the final price shown to you. If the airline website senses you are booking from a lower middle-class income zone, it would entice you with a cheaper flight price compared to if you booked the same flight from an upper-class income residential zone.
In short, you could be sitting next to someone who has paid half the cost of the ticket you are holding as you both head to the same destination.
Welcome to AI-Driven Analytics
Feel free to try the experiment; drop your friend with an Android device at the lower-income neighbourhoods as you drive off in the opposite direction towards Nairobi’s upper-income, leafy suburbs.
Armed with your Apple device, both of you try to book a flight to Kisumu simultaneously and let me know if you get the same price appearing on your device.
If that is too much work, a shorter experiment would be to book your next online taxi from your teenage son or daughter’s phone and compare the price displayed against the usual price you get charged for your regular trip to or from the office.
In both instances, there would be price differentials that consider personal data such as the device type, past expensive transactions you made without complaining, and the route you normally use.
Personalized Service – who gains?
Modern technologies such as Cloud Computing and AI have made it easy to collect personal data to provide personalized and convenient customer service. However, we often need to remember that convenience comes at a cost.
Most enterprises would use AI-driven models to mine your personal status knowledge to nudge that extra shilling out of your wallet. In most cases, you never know that you are paying different prices for the same products – depending on what category the algorithms decide you belong to.
Other than protesting as the Governor of Kisumu did, the Data Protection Act (2019) entitles you to be informed about how the AI-driven logic settled on a decision you may find uncomfortable to live with.
Transparency of AI models or ethical use of AI is increasingly important as society gets increasingly entangled with AI-driven logic.