Finding the Future
Escaping dataism
Hi all,
I wrote this up and hope you find it useful.
If neuroscientific research can be trusted, finding the future will require escaping the trap of linear, puzzle-solving thinking. To judge from numerous documented studies on patents and other signposts showing that innovation is slowing down, this thinking isn’t serving us. Compared to much of the twentieth century, the “exponentially progressing” twenty-first century is downright stagnant. And as Vaclav Smil has pointed out, the single decade of the 1880s accounts for more of the fundamental inventions and innovations making the modern world than any other before, or since. In our century, crunching huge datasets with powerful computers has been close to the only game in town—hardly real innovation (let alone invention). How do we restart innovation? How do we resuscitate culture?
Even if we take lateralization research as a metaphor, lessons are clear enough. We need to recover harmony with our “STEM” and “Liberal Arts” selves, we might say, as the right brain is more attuned to wonder rather than certainty, to music, language (it has no speech, but understands much about language), painting, and other venerable facets of rich cultural times like the European Renaissance. A sense of the fullness of the natural world and the human person, and the full range of faculties marked that time. A sense of the significance of the past did too.
Given the dominant cultural footprint of AI technology in consumer markets, business, academic research, healthcare, and government, it’s imperative that we liberate “AI” from “data science.” There is perhaps no more pernicious belief today than that induction—mechanical inference tied to prior observations or datasets—suffices for intelligence and that human smarts are a pale and slower version of induction, or “number crunching.” This hamstrings our ability to unlock the potential of our own minds, which use multiple forms of inference, including importantly abduction or commonsense reasoning, and it keeps us locked in a “puzzle world” where mechanical means seem proper for any end. It’s impossible to foster a thriving human-centered culture this way. Such a move would likely also prove a boon to AI, which has always been a diverse field until recently, and ideally encompasses cognate fields like neuroscience, cognitive psychology, and of course math and statistics.
We need paths forward, to the recovery of culture and innovation.
I believe we must take seriously relevant, latest scientific research, and take seriously the lessons of the past. Tall order.
Erik J. Larson