The Real Explanation of How AI Works

As the Information Age shifts to the Age of AI, it leaves many of us futurecasting what’s in store, particularly in recruiting and job search. Depending on who you ask, you’ll receive differing opinions on what AI truly is and isn’t. According to the Whitehouse’s Preparing for the Future of Artificial Intelligence report:

“There is no single definition of AI that is universally accepted by practitioners. Some define AI loosely as a computerized system that exhibits behavior that is commonly thought of as requiring intelligence. Others define AI as a system capable of rationally solving complex problems or taking appropriate actions to achieve its goals in whatever real world circumstances it encounters.”

Many new AI-inspired tools are hitting the talent acquisition market in an attempt to solve the significant challenges revolved around sourcing and attracting quality candidates. The hiring process is cumbersome and recruiters can feel overwhelmed with an overabundance of candidate submissions and the communication needed to source, screen and engage them. Therefore, recruiters are forced to speed up the process by investing in tools to automate many of their repetitive tasks. Sending out follow-up emails and scheduling interviews can all be automated, but just because something is being done faster, doesn’t mean it’s being done right.

Enter: AI. The one thing we can promise you it isn’t, is glorified automation. To best understand how AI works, you must acknowledge that experts refer to two variations of AI as either being weak AI or strong AI. There are weaker forms of AI, present in our everyday lives. The big data and algorithms used in financial market analysis and even on your Facebook feed to prioritize content you prefer is merely a slight introduction to AI at its most basic level.

As companies innovate to bring us stronger examples of AI, none have come quite as close as IBM’s Watson. Although it’s still not considered to fully be known as Strong AI, it does fall into the middle of the AI spectrum. This is because true strong AI is considered to be of the highest levels of intelligence, high enough to attain self-awareness and consciousness. Since AI has such a large spectrum, from a Facebook feed prioritizing your content to robots taking over the world and ending our existence, you can imagine how difficult categorizing AI can truly be. The common denominator is that AI learns from tasks, and does not just automate them.

Strong AI’s goal is to develop artificial intelligence that is functionally equal to a human’s. However, some of the most simple human tasks can be the most complicated for AI to tackle. Things like reading body language or identifying sarcasm that make humans the smartest machines of all, and we’re just not there yet with AI. Known as recursive self-improvement, true strong AI has the ability to learn without being taught.

In recruiting, many solutions are entering the market with AI capabilities. Some of these tools have the ability to conduct ongoing conversations with candidates to assess their personality traits, gauge their engagement and interest and determine if they’re qualified. These tools then are also to communicate to the recruiter which candidates are the most qualified which helps remove bias from the process, improve time-to-fill rates and reduce the recruiter’s workload so they can focus on their highest priorities. Other AI recruitment tools are able to score and rank or shortlist candidates from submitted information, but the conversation aspect is lost. Only one solution to date provides both the conversational functions and the scoring and ranking functions making it the most developed form of AI recruiting on the market! This tool is known as Karen and is powered by IBM’s very own Watson.

IBM’s Watson is able to comprehend concepts, text and materials through training via human experts to build what’s called a “corpus of knowledge” to gain literacy in the domain. The information is ingested by Watson to build indices and metadata that makes the information more accurate. Think of this as a knowledge graph that represents and leverages key concepts and relationships of the particular domain Watson is trained on. After the information is cultivated, Watson learns how to interpret the information through machine learning techniques in the form of questions and answers. Eventually, Watson is able to provide evidence-backed responses to questions it’s been trained on allowing it to enhance human expertise. Watson gets smarter the more you use it, and every user’s Watson is different depending on what the human user has taught it through action and processes.

Related Articles

Artificial Medical Products That Will Make Big Strides Towards Saving Lives in 2017

Stanford 100 Year Study Says Robots Won’t Try to Kill Us

Add Comment

What Baseball Means for Sales: Getting the Pitch Over the Plate
Four Ways Small Companies Can Frustrate The Big Boys
Traditional Artisanship Meets Technological Innovation at Mon Purse
10 Steps to Success that Will Always Work
“Disruption” Sounds Great….Until Your Life Is On The Line
NutraMilk: A Revolutionary Method of Preparing Healthy Food
How AI Solves the ATS Black Hole Problem for Recruiters
Pricing Jobs, Not People
Luxury and the Authentic African Safari: 10 Lodges That Do Both
A Beginner’s Traveling Guide to Abu Dhabi
Seven of the Most Unique Mini Golf Courses in America
Brent Bolthouse, the Mastermind behind Neon Carnival
The Top Ten Rated Minivans for 2017
The Top 10 Sports Cars To Look For In 2018
Ranking The Top 10 Luxury SUVs of 2017
The Top 10 Bugatti Models of All-Time
The 7 Best Men’s Watches for Under $500
The Seven Finest Moritz Grossmann Watches to Buy Right Now
The 10 Finest Watches From Baselworld 2017
The Top Five Omega Speedmaster Models of All-Time