Have you ever received a book recommendation that changed your life? Or discovered an old movie or music album that spoke to you in such a perfect way that it left you wondering how something so you was never suggested to you before? From a hole-in-the wall restaurant to a new podcast you listened to on your road trip; from a refreshing cocktail you’d never heard of before to a hike through the woods you never knew was there…
The quality of our leisure life — the space and time during which we’re free to choose the beat drumming our life’s symphony — stems from the quality of our discoveries; both of our own self, what we enjoy, and of the world providing countless wonderful possibilities.
And yet it can feel like a part-time job constantly encountering option overload and dealing with the paradox of choice. Although we live in a world of abundance, life satisfaction is less about quantity and more about quality. So how does one distill the myriad possibilities down to the best match for us right now? The answer could be: “recommendation engines”
Unfortunately, today’s recommendation engines, while ubiquitous in our lives, are fundamentally flawed and simply not as good as they should be. They’re fragmented, shallow, one-dimensional and biased… with limited understanding of the content they recommend, and even less understanding of the context in which they function. Adding injustice to injury, these algorithms are primarily driven by data that was snooped and stolen, tracked and taken without asking (Cf “Advertising Vs Recommendations”).
To the majority of today’s recommendation engines, users are not customers, they’re a commodity. The company-user relationship is defined by addictiveness and ad-sales, rather than quality of experience and personal wellbeing. Moreover, the algorithms are myopic and confined to only a sliver of your true self. They make assumptions based on what was rather than what might be, or who you wish to be.
Recommendations should help you grow into the person you want to become. At the very least, recommendations should meet you where you are, in your current mood, and enhance your experiences.
So of course, if a recommendation engine is fundamentally designed for misalignment with one person’s tastes, how bad can it be when it comes to finding the perfect gem for 2, 3 or 6 people? The time that two individuals can spend deciding what to watch or what to eat together can reach a ridiculous number of minutes or even hours. If it is already impossible to truly distill the uniqueness one user, it becomes simply absurd pretending to understand multiple sets of preferences at once — their contexts and how they align — without falling hastily into a series of stereotypes.
The passive, exploitative and manipulative relationship with our algorithms must evolve as we move into the next generation of artificial intelligence. We need an unbiased recommendation engine that knows us personally — our taste, interests, and desires. It’s the kind of relationship which requires a foundation in trust and transparency, with an agile adaptability which delivers what we want, when we want it.
All this embodies the “why”, why we’ve built Hai: the first pure, cross-domain and cross-platform recommendation engine. A personal assistant, companion and guide for making the most of your time.
Hai was designed to solve these problems of choice and attention by delivering customized recommendations uniquely tailored to the individual.
Built on your explicit input and feedback, Hai grows to understand the intangible spirit or essence that makes your clock tick. Instead of digging through archives or reading reviews online, Hai does all the work so you can spend less time searching and more time savoring the experience.
Rather than getting caught in filter and stereotyped bubbles of biased design (such as advertising), Hai helps you expand your horizons; to explore adjacent possibilities of your own taste and character. Hai finds the hidden gems, weaving through the mainstream and counter-stream to find your own special stream of conscious choice.
Free and accessible to anyone with access to the Internet, Hai democratizes the latest and greatest in Machine Learning. Powered by Crossing Minds algorithms, Hai is build upon the latest breakthrough in Collaborative Filtering, Deep Content Extraction and Semantic Graph Embeddings. Hai grows with, through, and alongside each person’s character and development to become a uniquely personal artificial intelligence. An AI that really understands you.
Platform agnostic, Hai connects with all your favorite platforms for seamless synchronization. It’s fluid and frictionless — how technology was meant to fit into our lives. Hai is not a content provider, nor an advertising platform, its incentives are fully aligned with the user.
We believe Hai is a true milestone in the recommendation space, at the beginning of an intimate, lifelong companionship between humanity and AI.
In fact, our long-term vision of Hai is an operating system, much like Samantha from the film Her. An intelligent presence that understands you and cares for you, and has your best interests at the heart of its code.
As the savvy Enlightenment figurehead Goethe put it: “Every day we should hear at least one little song, read one good poem, see one exquisite picture, and, if possible, speak a few sensible words.” Thriving, living life as our best selves, becomes a lot easier when we are regularly inspired by great art and culture. And that’s our vision for life with Hai: self-empowerment through experiences of awe and joy, on-demand. Curate your culture, feel the music in everyday life, and dance.
Hai is your companion and concierge, a muse.
It is with all those considerations at heart that Dr. Emile Contal, Alexandre Robicquet and Dr Sebastian Thrun co-founded Crossing Minds. They have spent their careers seeking to develop contextually-relevant, optimized recommender systems to deliver the best possible personalized recommendations for users at the most opportune time. Through user-filtering, item-filtering, content-filtering, deep learning, and transfer learning, the recommendation systems developed by the company seek to surface the best recommendations unique to each user, not simply for “users like you”. That’s the company promise.
To check it out, you can sign-up here.
Thank you for reading.