Are We Nearing the End of Human Innovation?

Press enter to search
Close search
Open Menu

Are We Nearing the End of Human Innovation?

By Jordan Speer - 05/05/2017

Lately, I’ve been viewing everything through the lens of deep learning. I’ve been asking myself, how will artificial intelligence (AI) change this job, or relationship, or business, or activity, or health crisis, or political situation, or [insert anything here]?

This line of thinking has been particularly relevant as we’ve been digging into the finer points of this year’s group of Apparel Top Innovator Award winners. For some, AI is the key enabler of the innovation on display. Consider EVISU, a lifestyle brand that has made incredible strides in digital marketing by using a marketing platform powered by AI that can test a virtually limitless combination of ad graphics and text across channels. The system discards what fails to capture consumer attention and keeps what works, learning as it goes — no human necessary.

This is one of several clear-cut examples. But I would venture to say that most of our Innovators are somewhere on the deep-learning trajectory. A great many have moved closer to the holy grail of retail: giving the customer what they want, when, where and how they want it and at the right (and profitable) price point by automating processes across the supply chain. Each is a unique variation, but the core theme is the same: put one version of data in one secure repository, accessible from anywhere, with appropriate transparency and visibility for all, eliminating redundancies, inefficiencies and human error and speeding the product lifecycle. Free up time for humans to do more creative, or value-added, or strategic-thinking things, such as developing a new performance technology, or assisting clients on the sales floor, or collaborating with another company on a unique partnership. This is fantastic progress.

Automation is not the same as AI, but it takes a company one step closer. Because once you have all that data in one place, clean and accessible, you can start to feed it into algorithms. Hello AI. The difference between AI and what we now call deep learning is the exponential quality of it.

We will one day soon have at our disposal the ability to recognize patterns in ways that are hard to fathom today. It’s already begun, but it’s only just started. Deep learning enables computers to recognize, for example, an irregularity so slight, that a human could not possibly have uncovered it, for not being able to process the billions of pieces of data that would have been required to ferret it out. This capability will alert us well in advance of pending problems, allowing us to prevent everything from illnesses to transportation failures to security breaches. Just imagine what deep learning can do for recognizing customer preferences, and the resulting opportunities for personalization and customization.

I listened to a TED Talk with Jeremy Howard last month entitled, “Will Artificial Intelligence Be the Last Human Invention?” I don’t have space to do it justice, but in a nutshell, what he says is that, unlike with previous enormous shifts in human history, such as from hunting and gathering to domestication, or animal energy to mechanical energy, the process of replacing human intelligence with artificial intelligence will not result in new types of jobs for humans. Deep learning will simply cut us out of the equation, because it has infinite opportunities for expanding on its own intellectual capabilities.

In just the past two years since Howard gave his first TED Talk about deep learning, computers have surpassed humans in their ability to recognize what’s in photos, and in their ability to understand English and Chinese. Amazingly, “there are now deep learning algorithms that are better at building the neural networks that create deep learning algorithms than humans are,” he says. There are areas today where we can say computers are still dumb, says Howard, “but in five years they won’t be.”

We often say that necessity is the mother of invention, but advances in deep learning may turn this equation on its head. Innovator CareZips is making life better by creating garments that a caregiver can more easily help a patient in and out of. This eliminates physical strain for the former, and humiliation for the latter. But what will it mean when that caregiver is a humanoid who possesses (or seems to possess) self-awareness due to deep learning of billions of data points that build emotional capabilities? Where will the real humans be in this scenario? It’s something to think about.

Today’s innovations are making a better world for tomorrow, but that could change. Tomorrow’s helpful innovation will harness technology without writing us out of the picture. We should weigh all innovations not merely on a scale of how they eliminate the hassles of life, but also on a scale of how they enrich it.

Jordan K. Speer is editor in chief of Apparel Magazine. She is waiting for a robot to take her job.

More Blog Posts In This Series

The Limitations of Repetition: A Breeding Ground for Innovation

Sometimes innovation comes from outside, in a flash, but more often than not it's the result of seeing things differently, after many years of accumulated experience and knowledge.

The Uncommon Good: Making Apparel in the 21st Century

If you’ve been watching the apparel industry for a long time, there are things you’ve likely come to realize about it. One of them? The inexorable march of time does not care about your apparel business model.

Breaking Down the Robot Takeover

If we leave robotic tasks to the robots, seeking to blend the best of what both humans and technology have to offer, we’ll reap the benefits of unleashing the creative forces of the human mind.