Phil Lawson: If I may, for a sec, compare the culture being developed right now in AI to the culture of a nation. Note what this means to a nation via the following excerpts from the book, Cultural Imperative: Global Trends in the 21st Century, by Richard D. Lewis:
“It [a nation’s culture] is an all-embracing pattern of a group’s entire way of life, including a shared system of values, social meanings, and agendas … Some of these attributes are subject to change, but the cultural framework generally endures … “
So far, 29-year-old white males (@ Google) represent the “majority” defining the culture in search and AI. It is from this perspective that search criterion and ranking is being established — that the cultural framework for AI is taking shape, that the patterns and ‘systems of value, the social meanings and agendas’ are forming. Therefore, we’d like to suggest (insist, really) that:
- Those designing (and socializing) AI must have a broad understanding of humanity (not just code / algorithms)
- And they must rethink omniscient programming approaches for search and AI / personal assistants
We know this: There is no magic algorithm that produces perfect search returns to a user. Human coders working for corporations are using editorial judgements to offer this service. Note the following:
“Humans are the ones who decide,” says a white paper commissioned by Google in August of 2012, page 11, “how the algorithm should predict the likely usefulness of a Web page to the user. These human editorial judgments are responsible for producing the speech displayed by a search engine … criterion for ranking search results … and … measure of a site’s value. Search engine results are thus the speech of the corporation.”
Certainly, there’s benefit in general universal search returns on our smartphones. But, when it comes to our own individual needs, our unique and evolving circumstances, when it comes to what makes us human in human situations, the information we require must reflect — it must speak to who we are, in each of our moments, all of which is subject to ongoing change as we shift and grow.
To meet this challenge, AI coders must transform omniscient-style programming approaches to include AI to human collaborations
that learn the human, interact with the human, and directly respond to his/her individual needs in real-time.