KITT in Knight Rider, Iron Man’s J.A.R.V.I.S, C-3PO from a galaxy far, far away — there’s a long-established fascination with meaningful machine-human interaction. This dynamic has characterized many of the recent developments in technology — in particular, those related to:
Artificial Intelligence (AI).
Machine Learning (ML), a method for achieving AI.
Deep Learning (DL), a specialized subfield of ML.
AI and ML capabilities have advanced exponentially in recent years, blurring the line between fantasy and reality and creating an unparalleled market opportunity for whoever can bring the technology to eager consumers. When lifeBEAM was seeking funding to launch their AI-enabled headphones, Vi, they turned to Kickstarter, literally capitalizing on the public appetite for “smart” products. They achieved their $100k funding goal in a mere 90 minutes, bringing in almost $1.7 million in donations by the conclusion of the campaign.
Self-driving cars, arguably the application most within reach for widespread use, are another prime example. This area is poised for extraordinary growth. 10 million self-driving cars are predicted to be on the road by 2020, and entrenched manufacturers like BMW and Mercedes, disruptive startups like Tesla, and even tech giants like Google are all rolling out self-driving models in the hopes of capturing future market share.
From autonomous cars to professional services to human genomics, companies are racing to perfect solutions and secure the first-mover advantage. In short, we’re experiencing an AI gold rush — and like many California 49ers discovered nearly two centuries ago, not everyone will cash in.
Successful AI products must be driven by one or both of two fundamental goals: aiding people in accessing and processing information and facilitating decision making. The crucial characteristic shared by both is that AI is treated as technology to supplement, not replace, human capability. Developers and applications that fail to grasp the continued necessity of human input are doomed to fail, while those that do are poised for sustainable success.
Recruiting is a prime example. Much has been made of the transformational impact that AI will have in recruiting; however, as any HR professional will tell you, most of these “transformative” applications lack the necessary capabilities to effectively recruit. As sophisticated and complex as AI has become in recent years, it cannot account for the uniquely human eccentricities and instincts involved in the recruitment process. On paper, transitioning the sourcing of talent to algorithms is an easy way to increase efficiency and accuracy. In reality, however, execution falls short of the mark.
An effective application, on the other hand, lies in the field of cancer research. There’s a tremendous wealth of research on the disease, with new data and resources introduced each day. As accurate diagnosis and treatment often fall on the shoulders of a single practitioner. It’s vital, though unrealistic, that they be cognizant of all available information. Combined with cancer’s highly individualistic nature (meaning every patient’s diagnosis is singular, and a treatment that proves effective for one is not guaranteed to work for another), the odds grow even grimmer.
As part of a tandem with a human counterpart, the artificial supplement bolsters the genuine intelligence (that is to say, human intelligence), propelling achievements to unrivaled heights, and bringing new hope to a diagnosis that can seem insurmountable.
The potential impact of the harmonious partnership between medicine and AI has not gone unnoticed. Technology brands and networks we engage with every day are diving into the fray: IBM Watson for Genomics, Google DeepMind, and the DREAM Challenges, are just a few examples. Facilitating how we process, sort, and analyze the data of disease will provide an invaluable advantage to modern medicine.
The most effective applications aren’t exclusive to matters of life and death, however. Google Maps, an application you likely use daily, encapsulates how AI and Machine Learning can impact our basic decision making. Determining the fastest route, indicating where to turn, advising against roads due to traffic conditions — these are the types of minor decisions that effective application can take out of our hands, in a similar, though more trivial manner to the diagnostic process.
These applications embody what should be the ultimate goal of each tech giant and nimble start-up: symbiotic human/machine processes and a co-mingling of artificial and human intelligence to the benefit of all.
If bringing AI to consumers is the “gold,” then DL is viewed by many as the ideal tool for extraction. It’s a term that’s dominated tech headlines over the past year, with applications both flashy (defeating human champions in Go) and business-oriented (enabling autonomous driving). This specialized field of ML facilitates the tailoring of AI to address specific human and business challenges, creating the monetizable link between technology and consumer. These recent, high-profile breakthroughs in DL are what’s powering much of the steeply accelerated investment in AI, as once-intangible potential begins to take shape.
While the bull market on AI is the result of man-generated interest, a “natural selection” process will determine the ultimate survivors. Divergent dynamics will eventually converge, gravitating towards the most effective business and interaction models and establishing the early standards for AI and ML technology.
2016 saw the start of the AI gold rush, but 2017 will be the year we begin dividing participants into winners and losers. Everybody’s playing, but only some will strike it rich.