6 min read
Opinions expressed by Entrepreneur contributors are their own.
Ever since the 1950s Turing test, machines have been in constant competition to outsmart their human counterparts. Alan Turing, way ahead of his time, wondered if machines could think like humans — a proposition that seemed unimaginable at the time. Now a part of our daily lives, artificial intelligence (AI) has transformed technology and the role of people in the workforce.
AI’s newest breakthrough is computer vision. This human-imitating technology actually appeared back in the 1970s in its earliest form, but is now transforming industries with its newfound capabilities. In fact, you most likely encounter computer vision multiple times a day without even realizing it. From unlocking your iPhone with facial recognition to Tesla’s autopilot driving function, computer vision is all around us.
This remarkable field of computer science attempts to replicate the complex mechanisms of human vision, allowing computers to rapidly identify objects and automatically react or make recommendations based on those images.
In the past, computer vision had not yet met its full potential, lacking the massive amount of visual data and deep learning algorithms we have today. Computer vision has made an exponential leap in recent years because of the enormity of digital images and video data that we now have access to. The technology’s recent boom has also heavily relied on advancements in neuroscience and research of neural networks. In order to replicate human vision with all its intricacies, the machine-learning algorithms had to be written to mimic the same neural paths our brain uses when viewing and processing images.
Related: When Should You Not Invest in AI?
Surpassing human capabilities
Computer vision is now even surpassing human capabilities, increasing from 50 percent accuracy a decade ago to 99 percent accuracy today. Technology even has a one-up on human vision because people suffer from biases when processing images and, unlike machines, we get tired. But with its potentially enormous commercial value, can machine vision truly replace human eyes across so many industries?
The agricultural industry has made a sharp shift toward high-tech, increasing all-around productivity by utilizing AI. Using aerial imagery, growers and farmers can gain instant insights in order to streamline operations and increase gains. Some agtech startups use computer vision to detect defects of a commodity for grading, which reduces post-harvest losses.
Access to this data boosts efficiency across the agricultural supply chain and helps achieve greater returns year-round. It would be impossible to procure this amount of data manually, but the success of this type of product is also heavily dependent on people. Farmers, traders and food enterprises are all still essential players in the growth of the agricultural industry. Also, while AI technology is extremely advanced, there are still flaws, and it is critical to have people to take care of machine errors. Even as machines take on human tasks, they will still need us to guide them through these tasks for the foreseeable future.
AI’s transformation of retail, automating the smallest processes
And that aspect of AI — that it can’t go it alone without our help thus far — is consistent across industries. Computer vision has also transformed retail, transitioning old-school brick-and-mortar into the digital age. AI technology is helping retailers improve customer service and create a personalized shopping experience.
On the grocery store front, some startups have even developed automated check-out systems that mount directly on top of existing carts and baskets. Unlike the latest barcode-scanning checkout systems, computer vision alone can identify each item in the cart, improving the customer experience and cutting costs. Implementing this technology is drastically more affordable than other installments and involves very little changes to the shop’s environment, yet it makes shopping much more convenient.
Related: The Highs and Lows of AI Based Ecosystem for Retailers
That being said, in-store retail is still up against the convenience of online shopping, and in-person stores will need to continue to cultivate their competitive advantage — human interaction. In fact, the reason some shoppers prefer brick-and-mortar stores is because they still seek a more human shopping experience.
So while the convenience of AI products will attract more customers, store attendants who can help customers on-site are still necessary for business. You can’t ask a device to physically show you where an item is located in the store.
The options for streamlining retail with machine vision go beyond in-store customer experience. AI technologies are automating even the smallest processes that were previously done manually, freeing consumers from having to go to a brick-and-mortar store and wait in line for every little thing. Now, there are even AI-based assessment technologies for smartphones, enabling users to objectively self-grade their device’s value for remote trade-ins. The yearly market for trade-ins is huge, with $2 billion returned to customers for trade-ins in 2020.
Related: How AI Is Revolutionizing Small Businesses
Computer vision in medicine: saving lives
In the field of medicine, computer vision is not just improving efficiency — it’s saving lives. The advent of machine-learning technology has transformed healthcare. From waiting rooms to emergency rooms, this technology is helping doctors diagnose, perform surgeries and administer care with information that could easily be missed by the human eye. In industries such as medical care that rely on human interaction, human biases can also sometimes affect patient treatment. Therefore, automated care can be a solution to confront widespread disparity and variability in the way patients are treated. This exemplifies the way in which AI technology and people need to work together in order to achieve the best results.
As vast amounts of data continue to flood our online world and we learn more about the science behind the human eye, computer vision is set to become nearly human — or in some cases, better. Yet experts warn that there are limits in neuroscience, and trying to map neural networks, let alone having a computer imitate them, is complex and bound to be imperfect. What’s more, computers lack the common sense and background knowledge that humans have. Machines may be less prone to errors, but relying on machines alone is a mistake. There is still great value in human intuition, and until machines are able to match that, the two will have to continue to work hand-in-hand.