Shouldn’t we aim higher than humanlike intelligence? Artificial AI to the rescue!

In difficult times, everyone seeks a panacea. “Technology will save us!” And with recent high profile Artificial Intelligence (AI) successes like AlphaGo and Watson, who wouldn’t want to jump on the AI bandwagon? The truth is that while AI research activity is picking up, advancements in AI capabilities are slowing down. The simple reason for this is that our best model for intelligence is the human mind, but the complexity of our evolutionary-based “Brain OS” is well beyond our current understanding, and we are now pushing up against the challenge of getting machines to exhibit the most complex of human cognitive abilities, such as abstract reasoning, creativity, and the application of world knowledge to problem solving.

Though the pursuit of machine-based humanlike intelligence could lead to transformative capabilities and commercial opportunities, not to mention widespread unemployment, it turns out we actually have humanlike intelligence today — in the form of humans. Indeed, creative human minds will ultimately develop the next generation of AI capabilities. But we have serious problems today that we can’t seem to solve using our human minds. So aren’t we seeking more than humanlike intelligence? Don’t we need superhuman intelligence? But how can we achieve that if AI is still trying to be more like a human, and we humans are, well, stuck being human?

Today there is an opportunity to develop superhuman intelligence by leveraging the complementary abilities of human cognition with the best available AI methods to create hybrid distributed intelligent systems. In other words, it is in our reach to build networks of humans and machines that sense, think, and act collectively with greater efficacy than either humans or machines alone. The emerging subfield of AI known as “human computation” is exploring exactly those opportunities by inserting humans into the loop in various information processing systems to perform the tasks that exceed the abilities of machine AI. For this reason, human computation is jokingly referred to as “Artificial AI”.

While AI has received much media and VC attention recently, Human Computation is the low-hanging fruit for achieving transformative capabilities in the near term. Wikipedia, the go-to resource for basic information about all things, is an early example of what can happen when the right machine-based coordination system interacts with networks of humans to collect and organize information. Wikipedia is an ingenius and pioneering example of human computation, but only scratches the surface of what is possible.

Should we continue to pursue traditional AI? Of course! But when it comes to solving the problems of today, shouldn’t we put a bit more effort into realizing the tremendous potential of our existing platform of 7.5 billion human-based processors? And in doing so, wouldn’t we potentially solve the labor problem by moving toward an economy that treats human cognition as a valuable resource? Frankly, if we are to believe that intelligent machines will someday take over the world, I would feel more comfortable if we kept a few humans in the loop.