I was recently asked if, because Zilliant uses artificial intelligence (AI), we are seeing a reduced need for technical consultants and data scientists to a) be involved in customer deployments and b) stay involved through the ongoing use of our solution. The short answer? Absolutely not! For the same reasons we see Amazon, IBM and Facebook clamoring to hire talent, our in-house experts are as critical as ever. Let me dig into why...
Designing AI that works is not a simple task. Deploying AI that's embraced and acted upon by humans is even more difficult. Generally, I perceive three common misconceptions about AI:
- It turns on with the flip of a switch.
- It's black box.
- It will replace us all.
These couldn't be farther from the truth.
First Misconception: Flip a Switch to Turn on AI
Imagine you hire a new employee whom, based on their professional and educational credentials, is a genius and topical expert. How would you approach their first month on the job? Would you give them their badge, desk, computer, and leave them be without instruction, expecting they immediately know what to do and how to do it? Or would you have a plan in place to bring them up to speed and manage their performance to help them succeed and have maximum impact in their role?
Now, what if it was a machine instead of a human, would you flip a switch and walk away? Of course not. Even if the machine is using unsupervised learning, someone still needs to supervise the machine and its outputs. AI is not a plug-and-play proposition.
At Zilliant, we define AI as the capability of a machine to imitate intelligent human behavior. And each AI software is intended to imitate specific behavior.
AI is not created equal, nor is it designed for homogenous use cases. If you asked Zilliant IQ to safely drive your car between locations, beat a world champion at their game, or to rearrange your calendar, you will not be happy with the result. Today, for AI to be effective and useful, it needs to be trained on the right set of data - and lots of it - and for the right desired outcomes. To use a bad "Lord of the Rings" analogy, there is no "one ring to rule them all" in the world of AI. Evaluate AI solutions based on the specific business problems you’re trying to solve.
Second Misconception: AI is Black Box
Not all AI is black box, although AI sometimes can be black box. For instance, deep learning is inherently black box as there’s no means to look inside a machine’s neural network and make sense of it. In business, however, there are many interactions where the answer “because the machine says so” will be an unacceptable answer.
Zilliant’s AI software is designed to think and act like the most intelligent and experienced sales rep, backed by insights from the business’ best analysts. And not just any business, but the specific business, business unit, regional unit, product unit, etc. that we're currently solving for. When we work with customers, our dedicated project teams become experts in their businesses so that we can, in return, design solutions that deliver value by, at least, an order of magnitude greater than what they had before.
By working together with our customers, we can fine-tune our AI-driven platform to address the factors relevant to each customers' unique business challenges. Taking this approach, putting experts in the driver's seat of an AI-driven solution, delivers a two-fold benefit. First, our customers' in-house experts can pull unique profit and revenue levers to alter guidance and drive execution based on top-level business objectives or swift market changes. Second, by deploying AI guidance that's highly relevant to the realities of each unique business, the end-users more readily adopt, act on, and embrace the output.
Third Misconception: AI Will Replace Us All
Finally, AI is not going to replace us all. Where a machine can do a job better and faster, yes, we will see human displacement in jobs and functions. But the biggest opportunity today, and what companies should be looking for, is empowering employees and organizations to perform better by augmenting and complementing their actions with AI solutions.
In our case, we help companies achieve revenue and profit goals by knowing more about their customers than possible with any number of human analysts. From that knowledge, we deliver actionable insights about those customers to sales teams that translate to immediate revenue growth, greater profit, and happier, more loyal customers. Our solution doesn’t replace sales reps or analysts; it makes them a whole lot better.
Successfully Deploying AI Software
If you want to read another pragmatic outlook on AI in business, I enjoyed a recent opinion post along this same vein on Information-Management.com: “Predictions 2018: AI is tough stuff and many organizations will fail at it." When evaluating AI, B2B leaders need to understand the problem they’re trying to solve for, and then match a company and a solution that can help. The first thing we do with any potential customer is a consultative, diagnostic engagement. This allows both parties to identify potential problems, and to assess whether our AI solutions are the right fit for the business. If you’d like more information, join us for a webinar.
Finally, data scientists and technical consultants need not fear the machine. In fact, companies will need more of them to create AI software solutions that work in B2B organizations. And if you’re a data scientist or technical consultant, we’re hiring!
About the AuthorFollow on Twitter Follow on Linkedin