.png)
https://ift.tt/2ry2iv2
Artificial intelligence-powered automation is the name of the game in 2018: It’s being used to disrupt recruiting, medicine, customer service and numerous other industries across the board. This new force in the business world has resulted in some truly transformative technologies, but it’s also resulted in confusion and unwarranted applications of AI.
A business’s worst fear is lagging behind and becoming irrelevant. I've found this fear has spurred executives to latch onto AI and automation without fully understanding both its practical applications and its limitations. The triumphs of machine learning in products such as Uber’s ride ETA or Facebook’s facial recognition feature have overshadowed failures such as Microsoft’s Tay, Facebook's chatbots developing their own language, or Amazon's Alexa suggesting inappropriate content to a child.
As these failures demonstrate, though, AI has limitations and is not suited to every product or business. Here are what I consider the top three limitations of AI in 2018:
1. Automation is labor-intensive and full of hidden costs.
Successful automation unavoidably begins with humans training algorithms through a labor-intensive data labeling process. For instance, in the supervised learning process used to train self-driving cars, autonomous vehicle companies hire hundreds of people to manually annotate hours of video feeds from prototype vehicles. Manual data labeling is also being used by Facebook, which employs over 7,000 “content reviewers,” who flag content with the goal of giving an algorithm an ever-increasing corpus of data to learn from. And when the team at my company was developing the first version of our matching algorithm, it required 8,000 hours of manual labor in the form of expertise categorization to build a working prototype.
But even after the algorithm is developed, the need for human labor doesn’t disappear. For instance, the Facebook content-flagging algorithm is already built, but the need for content reviewers will continue until the algorithm performs as well or better than humans. There’s a continual push-pull between humans and algorithms that continues for years before a process becomes truly, fully automated.
For some businesses, this labor can be outsourced, minimizing the associated costs. For others, the investment in hiring skilled stateside employees to train the algorithm is worth it. The important thing is that you realize what you’re getting into, and examine the true cost of automation before investing in it.
FINANCE
FINANCE
via Forbes - Entrepreneurs https://ift.tt/dTEDZf
May 11, 2018 at 03:31PM
No comments:
Post a Comment