There’s pleasure to be had in the hunt for gratification for sure. That’s why we love Tinder, flash sale sites, and fishing (like the saying goes, “they call it fishing, not catching.”)
But when you don’t want to hunt—but just to find–search technology is outmoded and stuck in the ’90s, giving us clunky, cluttered, and impersonal lists we have to wade through. What most of us are doing when we “search” is not enjoying the process of “searching” but doing what we have to do to get what we want. Search is a means to an end, not an end in and of itself. And as our expectations for precise “rightness for me” increase, and the amount of data to sift through and the velocity with which that data changes increases as well, search is more and more unsatisfactory.
This is where AI comes in. What we want (maybe subconsciously and maybe uneasily) is some friendly, utterly trustworthy omniscient machine to pimp our life. We want the right thing at the right time, every time, without having to formulate the perfect search query. Filters help some by giving us lots of drop down menus to continually refine our search, but why doesn’t some bot already know that I always want the highest discount Prada bag that will fit my current laptop, has “feet” to avoid getting grimy, and is the edgiest model without being tacky? No filtering system can do this (other than Prada and highest discount, and it is amazing how many sites lack even that).
Since college, I’ve believed that there is some secret key to people’s preferences. My working theory for many years was that most of one’s food preferences could be predicted accurately from one key preference set – do you prefer Nilla Wafers or Oreos? I still think this is true, and you could cross reference that with something like, Hawaii or Costa Rica, or maybe book or movie, or Mercedes or BMW, or my latest — front license plate or risk a ticket because your car looks so much better without one. I’m actually shocked (and incredulous) that Amazon can’t (or won’t?) do this for me yet, given how many tells I’ve given them over the years. Perhaps they can but don’t because they fear I’d think it was creepy?
Which brings me to my next point. AI has to be paired with true trust to really work the way we, in our heart of hearts, want it to work. We need to be incentivized to both give our AI the biggest data set about us possible, even the most secret stuff, and to trust that what our AI tells us we want and need is not harmful.
There’s been a lot of writing about the trust economy. Throughout time, we’ve used various signals to suggest that we’ve made the right choice. We hire using educational institutions as proxies for potential employee value (the Ivy effect). We date based on who our prospective boyfriend dated before us. We recognize people of a similar social class by the clothing brands they wear and the cars they drive. For many years, we relied on big buildings (the “edifice complex”) or flashy brand spends to signal substantiality and permanence of companies where that was important (e.g., insurance or banking companies).
Now, we increasingly buy based on the reviews of our friends or our friends’ friends or even the aggregate of the masses (Uber or Airbnb ratings), so the edifice is being replaced with the social referent. But this level of sharing economy and social media generated trust isn’t sufficient to allow AI access to an adequate data set about each of us to fuel truly effective AI.
What we really need to achieve the nirvana of effortless gratification instead of endless searching and filtering is great AI plus some sort of blockchain of trust. How can we construct a completely secure and ethical repository of all of the data that AI needs to truly KNOW us? An AI that will fulfill our desires, but not exploit us? We need to trust it like we trusted our mothers when we were small, or even more so. Google’s “Do No Evil” promise, the promise of any other brand (the equivalent of the edifice effect) isn’t enough — we need another solution that transcends brands and even our conception of what a business enterprise or government can be.
This piece is a lot harder than AI in my opinion, and something that we haven’t even begun to tackle. I’m hoping we’ll see some sophisticated thinking in this area soon.