Quality is important when it comes to hiring candidates for a job, isn’t it? Then, it is also obvious that quality be important for those who are responsible for recruitment too. This means that if you are hiring some one to help you recruit candidates for your organization, you also need to make sure you are hiring the right people for the job. This is because one wrong interaction between your recruiting assistant and potential candidate can hamper the interview and have the candidate to drop off on his own. But today, most organizations are depending upon AI assistants to help with the job. This makes it even more important to evaluate the chatbot quality, one that can deliver the best possible candidate experience. Let’s help you understand the same with the help of a few examples.
Siri, with its limited communication options
Users use Siri for a variety of topics.
This is because Siri has an impressive knowledge about almost every topic,
considering the breadth of functions it is able to perform. However, this vast
functioning means that Siri knows a little about everything, but does not have
enough depth about each topic. Moreover, Siri has a very surface-level
understanding of the language humans use to communicate. This is why users need
to communicate with Siri in a very specific and limited manner. It’s just the
same like you landing in a foreign country where people understand only their
native language, but not English; and you know only the basics of the native
language. This hampers your natural way of communicating, which doesn’t make
you as comfortable as you should be.
Using this in the field of recruitment,
candidates would rather speak to someone who can understand their language and
intentions, rather than wasting their time trying to express something that chatbots
don’t understand much. Bots create a restrictive experience with Y/N questions,
or multiple-choice questions to candidates as communication options, which is
certainly not the best possible candidate experience. The natural language
processing of such bots is not sophisticated enough.
Poncho, with its set decision tree
When you visit a store, your experience
highly depends on the interaction you have with the attendant there. If the
attendant is attentive and understand your requirements, giving you a relevant
recommendation, you’ll be looking forward to shop from there again. But, what
if the attendant doesn’t seem to care to understand what you want, and presents
you with options that don’t match your preferences? You wouldn’t want to enter
the store again! Poncho is one such chatbot that was designed to tell people
about the weather. But, users found it difficult to explain their questions to
Poncho. Chatbots like Poncho rung on something called a decision tree, where it
becomes difficult to identify the question based on the question type. When a
chatbot isn’t good at fixing a mistake or doesn’t seem to understand your
intent, it’s a chatbot that is powered by a decision tree.
This isn’t something that recruiting
leaders would want in their recruiting chatbot. If a candidate deviates slightly
from the chatbot’s set path, the entire conversation will be derailed.
TayTweets, with its Machine Learning
capabilities
An AI is generally considered smart, but
not smart in all ways. That is because AI will quickly learn, but it doesn’t
know what to learn, and what not to learn. It doesn’t have a mind of its own,
so it just absorbs whatever you provide it with. You provide AI with data and a
framework for interpreting the data; and it simply absorbs all of it, with its
machine learning capacities. Thus, there is a lot that goes into following a
governance process to ensure that AI does not learn from any wrong data.
TayTweets is one such Microsoft-released AI that used machine learning for
communication. The machine learning algorithms used general unstructured data
that were fed to TayTweets by the Twitter community, but was not subjected to
screening by the Microsoft team. Within a short time, the AI was seen to ne
sending out derogatory and racist messages throughout the account! This is a
great example of understanding how everything can go wrong is you don’t have
the right governance around the data you are using to power machine learning.
A chatbot that lets in unstructured data
into its AI machine learning process can make things go extreme in the case of
recruiting!
Artificial Intelligence, or AI, is a
fascinating new technology that has been transforming recruitment currently.
But, there are lots of things to be kept in mind while using AI, as you can
learn from the examples above. You thus need to partner with leading talent
acquisition companies who pick an AI that creates a personal, intuitive, and
human experience for potential candidates throughout the recruitment process.
Genuine and knowledgeable talent acquisition companies in India, like WalkWaterTalent Advisors, would use conversational AI that have great communication and
understanding skills, rather than one that works off a decision tree.
Comments
Post a Comment