Por qué el auge de la IA hace que la inteligencia humana sea más valiosa que nunca
Why the Rise of AI Makes Human Intelligence More Valuable Than Ever
By Michael Dix
In the popular TV show Sherlock, visual depictions of our hero’s deductive reasoning often look like machine algorithms. And probably not by accident, given that this version of Conan Doyle’s detective processes tremendous amounts of observed data—the sort of minutiae that the average person tends to pass over or forget—more like a computer than a human.
Sherlock's intelligence is both strength and limitation. His way of thinking is often bounded by an inability to intuitively understand social and emotional contexts. The show's central premise is that Sherlock Holmes needs his friend John Watson to help him synthesize empirical data into human truth.
In Sherlock we see the analog for modern AI: highly performant learning machines that can achieve metacognitive results with the assistance of fully cognitive human partners. Machine intelligence does not by its nature make human intelligence obsolete. Quite the opposite, really—machines need human guidance.
And they need it now. The world is firmly in the throes of the Fourth Industrial Revolution, defined generally as the age of machine intelligence and cyber-physical systems. As in previous waves, the way that we work and value effort is changing dramatically, but this time the disruption is concentrated over one human generation instead of several. Repercussions in economic and political disruption are already being felt. Future shock is S.O.P.
In the age of bots and androids, we must be more human
How do workers, leaders and makers navigate this rapidly shifting landscape successfully? We start with what makes us human. As a species, Homo sapiens display three traits that differentiate us from all other life on the planet, and from our intelligent machines.
Ingenuity. We imagine and create entirely new things, from complex tools to art to smart devices. Creating and driving new business contexts requires imagination that even the most intelligent machine doesn’t possess. An AI can learn the game of Go so well it consistently beats the best human players, but unlike humans, that AI can’t invent a new game. And it can’t conceive a desire to learn about rockets.
Thanks to cloud computing, a garage entrepreneur can turn an idea into a minimum viable product in a few months. Automated manufacturing and 3D printing means that individuals can compete with larger companies by producing artisanal or novel items at scale with low go-to-market cost.
Adaptability. We survive by changing our ways to suit radically different environments and circumstances; this trait has been key to our success as a species. An intelligent assistant can improve its modeling of human interaction and return smarter results about likes and dislikes, but it isn’t an autonomous being. It cannot execute outside of its context. For example, my car’s bot can call for help if I’m stranded in a snowstorm, but it can’t choose to put chains on my tires.
Human workers are adapting to Fourth Wave production environments powered with smart devices and machine tools. Some people are optimizing skills or retraining to work with new devices or manage industrial robots. All over the world, we see people accepting the "new normal" of using online services for work and play.
Ethical capacity. We construct moral frameworks from which to evaluate courses of action, anticipate consequences, and make value judgments. Metacognition is our species’ unique superpower, the essence of autonomy. It enables us to perceive “good” and “evil,” to empathize, and to make ethical choices (or not). This particular trait allows humans to accomplish collective goals and thrive in very large social groups (most of the time). It’s critical to homo sapiens’ continued success in the time ahead.
Humans determine whether machine learning capabilities are used for good or ill. An AI can sift through data to help improve workplace safety and aid efficiency. It can draw insights to improve human learning, or find solutions to global health and safety problems. Or bots can be directed to manipulate data in support of shady operatives looking to subvert human laws and norms. An intelligent machine’s “moral code” is only as good as the definitional data it receives from humans; it doesn’t have metacognitive awareness.
Machines cannot remake human futures. That’s on us.
Human traits are necessary not just for training intelligent machines, but for redefining the work we humans want to do and how to value it. Carnegie-Mellon's Manuela Veloso recently coined the term "symbiotic autonomy" to describe the way in which humans and machines will be able to create new forms of work. She predicts that soon "it will be hard to distinguish human agency from automated assistance—but neither people nor software will be much use without the other."
Which sounds a lot like John Watson and his BFF Sherlock.
We need to bring our human abilities not only to engineering sophisticated algorithms but also to synthesizing machine learning insights into new human potential. Over the next five years, machine learning will become democratized. Intelligent services will be easier to integrate into all aspects of business, not just obvious areas like platforms. The economic disruption of cyber-physical machines displacing unskilled and low-skilled labor will continue for the foreseeable future.
Can productivity take new forms for people without the capability or means to retrain? People already provide enormous amounts of that data through apps and cloud services, including social media, so perhaps we develop a "work" model based on generating data for intelligent algorithms. So far, the value of individual input through normal cyber activity is largely uncompensated except in terms of diversion or access to information.
Intentional Futures sees the Fourth Wave as the point where the science of human learning becomes crucial to developing symbiotic autonomy in enterprise.
Intelligent technology can improve human outcomes, but it needs educated, imaginative humans to realize full viability. And we humans need to exercise our own capacity for ethics to adapt how we think about and value work. "Democratization in AI” is more than enterprise access to skills and data; it must include access to the contexts and training that strengthen human imagination.
Image Credit: Shutterstock
"Before co-founding Intentional Futures, Michael spent eight years at Microsoft leading product management and marketing strategy teams. He also worked in Beijing for News Corporation establishing one of the first Internet ventures in China."