The problem with Artificial Intelligence and Machine Learning, at the moment, is a problem of use, not of technology.
Companies mainly implement AI in their products and services to maximise economic results, and they fail (mostly) to actually deliver value to their users.
Take the advertising industry, for example. AI and ML are used by platforms to predict interests and needs based on data collected from your online behaviour. It is currently a widely inaccurate utilisation of the technology, that’s why you get exposed for months to ads from that site you visited once while you were researching your competitors for the next management meeting.
It is so because what currently matters is not that you get an ad that is relevant to you or that the advertisers message gets exposed to the correct audience. What matters, at this particular moment in history, is for the platforms to sell as many ads as they can. And since their competition is even less accurate or completely unmeasurable, they thrive with very little conversion rates while they make the rules they feel are more appropriate to achieve what they want.
Now, imagine a slightly different application of the same technology.
We have noticed that in past weeks, you have visited sites of car dealers. Would you like us to push some car offers from local dealers to your timeline?
You have visited this restaurant three times in the last month, would you like me to add it to your favourite restaurants in town? I could push some of their lunch offer to your inbox, if you want me to. Just tell me how frequently you’d like to receive them.
I see you’ve been at events about business and management in the past six months. Here are a bunch of groups you might be interested in. Also, there is a special deal on the Business and Management magazine if you subscribe by the end of the year. Do you want to take it?
Asking questions instead of assuming, is a great rule of thumb for interpersonal relationships. The same should be valid for interactions with a machine.